TA_Screening_0_59.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "purple-youth",
   "metadata": {},
   "outputs": [],
   "source": [
    "from pandas_datareader import data as pdr\n",
    "from yahoo_fin import stock_info as si\n",
    "import yfinance as yf\n",
    "import pandas as pd\n",
    "import datetime\n",
    "import time\n",
    "import talib\n",
    "from talib import *\n",
    "import matplotlib.pyplot as plt\n",
    "from xgboost import XGBClassifier\n",
    "import xgboost as xgb\n",
    "import numpy as np\n",
    "import pickle\n",
    "import os\n",
    "os.environ[\"PATH\"] += os.pathsep + r'C:\\Program Files\\Graphviz\\bin'\n",
    "yf.pdr_override()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "proper-locking",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "nuclear-diversity",
   "metadata": {},
   "outputs": [],
   "source": [
    "def add_indicators():\n",
    "    \n",
    "    \n",
    "    \n",
    "    upper_band, mid_band, lower_band = BBANDS(df['Adj Close'],timeperiod=slow, nbdevup=2, nbdevdn=2, matype=0)\n",
    "    \n",
    "    upper_band_nine, mid_band_nine, lower_band_nine = BBANDS(df['Adj Close'],timeperiod=fast,\n",
    "                                                             nbdevup=1.645, nbdevdn=1.645, matype=0)\n",
    "    \n",
    "    upper_band_one, mid_band_one, lower_band_one = BBANDS(df['Adj Close'],timeperiod=really_fast,\n",
    "                                                             nbdevup=1, nbdevdn=1, matype=0)\n",
    "    \n",
    "    upper_band_slow, mid_band_slow, lower_band_slow = BBANDS(df['Adj Close'],timeperiod=slow,\n",
    "                                                             nbdevup=1.645, nbdevdn=1.645, matype=0)\n",
    "    \n",
    "    d_ema = DEMA(df['Adj Close'], timeperiod=really_fast)\n",
    "    E_M_A = EMA(df['Adj Close'], timeperiod=fast)\n",
    "    ht_trend = HT_TRENDLINE(df['Adj Close'])\n",
    "    kama = KAMA(df['Adj Close'], timeperiod=fast)\n",
    "    ma_fast = MA(df['Adj Close'], timeperiod=fast, matype=0)\n",
    "    ma_really_fast = MA(df['Adj Close'], timeperiod=really_fast, matype=0)\n",
    "    ma_slow = MA(df['Adj Close'], timeperiod=slow, matype=0)\n",
    "    #mama, fama = MAMA(df['Adj Close'], fastlimit=really_fast, slowlimit=slow) < this gave me issues?\n",
    "    #mavp = MAVP(df['Adj Close'])\n",
    "    mid = MIDPOINT(df['Adj Close'], timeperiod=fast)\n",
    "    mid_price = MIDPRICE(df['High'], df['Low'], timeperiod=fast)\n",
    "    sar = SAR(df['High'], df['Low'], acceleration=.02, maximum=.2)\n",
    "    sarext = SAREXT(df['High'], df['Low'], startvalue=0, offsetonreverse=0, accelerationinitlong=.02, accelerationlong=.02, accelerationmaxlong=.2, accelerationinitshort=.02, accelerationshort=.02, accelerationmaxshort=.2)\n",
    "    sma = SMA(df['Adj Close'], timeperiod=slow)\n",
    "    tema = TEMA(df['Adj Close'], timeperiod=slow)\n",
    "    trima = TRIMA(df['Adj Close'], timeperiod=slow)\n",
    "    wma = WMA(df['Adj Close'], timeperiod=slow)\n",
    "\n",
    "    #this is some of the beginning stuff\n",
    "\n",
    "    O_B_V = OBV(df['Adj Close'], df['Volume'])\n",
    "    A_D_O_S_C = ADOSC(df['High'], df['Low'], df['Adj Close'], df['Volume'], fastperiod=fast, slowperiod=slow)\n",
    "    O_G_chaikin = AD(df['High'], df['Low'], df['Adj Close'], df['Volume'])\n",
    "    HT_DCper = HT_DCPERIOD(df['Adj Close'])\n",
    "    HT_DCphase = HT_DCPHASE(df['Adj Close'])\n",
    "    inphase, quad = HT_PHASOR(df['Adj Close'])\n",
    "    r_sin, leadsin = HT_SINE(df['Adj Close'])\n",
    "\n",
    "    #volatility\n",
    "    atr = ATR(df['High'], df['Low'], df['Adj Close'], timeperiod=slow)\n",
    "    natr = NATR(df['High'], df['Low'], df['Adj Close'], timeperiod=slow)\n",
    "    t_range = TRANGE(df['High'], df['Low'], df['Adj Close'])\n",
    "\n",
    "    #below here are momentum ind\n",
    "\n",
    "    adx = ADX(df['High'], df['Low'], df['Adj Close'], timeperiod=fast)\n",
    "    adxr = ADXR(df['High'], df['Low'], df['Adj Close'], timeperiod=fast)\n",
    "    apo = APO(df['Adj Close'], fastperiod=really_fast, slowperiod=fast, matype=0)\n",
    "    aroon_d, aroon_u = AROON(df['High'], df['Low'], timeperiod=fast)\n",
    "    aroon_osc = AROONOSC(df['High'], df['Low'], timeperiod=fast)\n",
    "    bop = BOP(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    cci = CCI(df['High'], df['Low'], df['Adj Close'], timeperiod=fast)\n",
    "    cmo = CMO(df['Adj Close'], timeperiod=fast)\n",
    "    dx = DX(df['High'], df['Low'], df['Adj Close'], timeperiod=fast)\n",
    "    macd, macdsig, macdhist = MACD(df['Adj Close'], fastperiod=fast, slowperiod=slow, signalperiod=really_fast)\n",
    "    macdex, macdexsig, macdexhist = MACDEXT(df['Adj Close'], fastperiod=fast, fastmatype=0, slowperiod=slow, slowmatype=0, signalperiod=really_fast, signalmatype=0)\n",
    "    macdfixd, macdfixdsig, macdfixdhist = MACDFIX(df['Adj Close'], signalperiod=really_fast)\n",
    "    # more momo's\n",
    "\n",
    "    mfi = MFI(df['High'], df['Low'], df['Adj Close'],df['Volume'],timeperiod=fast)\n",
    "    min_di = MINUS_DI(df['High'], df['Low'], df['Adj Close'], timeperiod=fast)\n",
    "    min_dm = MINUS_DM(df['High'], df['Low'], timeperiod=fast)\n",
    "    momo = MOM(df['Adj Close'], timeperiod=really_fast)\n",
    "    plus_di = PLUS_DI(df['High'], df['Low'], df['Adj Close'], timeperiod=fast)\n",
    "    plus_dm = PLUS_DM(df['High'], df['Low'], timeperiod=fast)\n",
    "    ppo = PPO(df['Adj Close'], fastperiod=really_fast, slowperiod=fast, matype=0)\n",
    "    roc = ROC(df['Adj Close'], timeperiod=fast)\n",
    "    rocp = ROCP(df['Adj Close'], timeperiod=fast)\n",
    "    rocr = ROCR(df['Adj Close'], timeperiod=fast)\n",
    "    rocr_hund = ROCR100(df['Adj Close'], timeperiod = fast)\n",
    "    rsi_fastk, rsi_fastd = STOCHRSI(df['Adj Close'], timeperiod=fast, fastk_period=slow, fastd_period=really_fast, fastd_matype=0)\n",
    "    trix = TRIX(df['Adj Close'], timeperiod=slow)\n",
    "    ult_osc = ULTOSC(df['High'], df['Low'], df['Adj Close'], timeperiod1=really_fast, timeperiod2=fast, timeperiod3=slow)\n",
    "\n",
    "\n",
    "    #old some of the first added\n",
    "    R_S_I = RSI(df['Adj Close'], timeperiod=slow)\n",
    "    slowk, slowd = STOCH(df['High'], df['Low'], df['Adj Close'], fastk_period=fast, slowk_period=slow, slowk_matype=0, slowd_period=slow, slowd_matype=0)\n",
    "    fastk, fastd = STOCHF(df['High'], df['Low'], df['Adj Close'], fastk_period=fast, fastd_period=really_fast, fastd_matype=0)\n",
    "\n",
    "    real = WILLR(df['High'], df['Low'], df['Adj Close'], timeperiod=slow)\n",
    "\n",
    "    # below are the TA indicators\n",
    "\n",
    "    two_crows = CDL2CROWS(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    three_crows = CDL3BLACKCROWS(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    three_inside = CDL3INSIDE(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    three_line = CDL3LINESTRIKE(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    three_out = CDL3OUTSIDE(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    three_stars = CDL3STARSINSOUTH(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    three_soldier = CDL3WHITESOLDIERS(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    baby = CDLABANDONEDBABY(df['Open'], df['High'], df['Low'], df['Adj Close'], penetration=0)\n",
    "    adv = CDLADVANCEBLOCK(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    belt_hold = CDLBELTHOLD(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    breakaway = CDLBREAKAWAY(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    closingmara = CDLCLOSINGMARUBOZU(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    baby_swallow = CDLCONCEALBABYSWALL(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "\n",
    "    #more TA\n",
    "\n",
    "    counter = CDLCOUNTERATTACK(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    dark_cloud = CDLDARKCLOUDCOVER(df['Open'], df['High'], df['Low'], df['Adj Close'], penetration=0)\n",
    "    doji = CDLDOJI(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    doji_star = CDLDOJISTAR(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    dragon_doji = CDLDRAGONFLYDOJI(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    engulf = CDLENGULFING(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    evening_star = CDLEVENINGSTAR(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    gapside = CDLGAPSIDESIDEWHITE(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    gravestone = CDLGRAVESTONEDOJI(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    hammer = CDLHAMMER(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    hang_man = CDLHANGINGMAN(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    harami = CDLHARAMI(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    harami_cross = CDLHARAMICROSS(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "\n",
    "    #more TA\n",
    "\n",
    "    high_wave = CDLHIGHWAVE(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    hikkake = CDLHIKKAKE(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    hikkake_mod = CDLHIKKAKEMOD(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    pidgeon = CDLHOMINGPIGEON(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    id_three_crows = CDLIDENTICAL3CROWS(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    in_neck = CDLINNECK(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    inv_hammer = CDLINVERTEDHAMMER(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    kicking = CDLKICKING(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    kicking_len = CDLKICKINGBYLENGTH(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    ladder_bot = CDLLADDERBOTTOM(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    doji_long = CDLLONGLEGGEDDOJI(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    long_line = CDLLONGLINE(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    marabozu = CDLMARUBOZU(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "\n",
    "    #more TA\n",
    "\n",
    "    match_glow = CDLMATCHINGLOW(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    mat_hold = CDLMATHOLD(df['Open'], df['High'], df['Low'], df['Adj Close'], penetration=0)\n",
    "    morning_doji = CDLMORNINGDOJISTAR(df['Open'], df['High'], df['Low'], df['Adj Close'], penetration=0)\n",
    "    morning_star = CDLMORNINGSTAR(df['Open'], df['High'], df['Low'], df['Adj Close'], penetration=0)\n",
    "    on_neck = CDLONNECK(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    pierce = CDLPIERCING(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    rickshaw = CDLRICKSHAWMAN(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    rise_fall = CDLRISEFALL3METHODS(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    sep_line = CDLSEPARATINGLINES(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    shooting_star = CDLSHOOTINGSTAR(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    sl_candle = CDLSHORTLINE(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    spin_top = CDLSPINNINGTOP(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    stalled = CDLSTALLEDPATTERN(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "\n",
    "    #more TA\n",
    "\n",
    "    stick_sand = CDLSTICKSANDWICH(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    takuri = CDLTAKURI(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    tasuki_gap = CDLTASUKIGAP(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    thrust = CDLTHRUSTING(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    tristar = CDLTRISTAR(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    three_river = CDLUNIQUE3RIVER(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    ud_two_gap = CDLUPSIDEGAP2CROWS(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "    down_three_gap = CDLXSIDEGAP3METHODS(df['Open'], df['High'], df['Low'], df['Adj Close'])\n",
    "\n",
    "    #76 vars\n",
    "\n",
    "    #are_all_zero = (test_TA == 0).all()\n",
    "    #true if all values are 0\n",
    "    #false if contain a non 0'''\n",
    "\n",
    "    #df.drop(['Close'], axis =1, inplace = True)\n",
    "\n",
    "    df['upper_band'] = upper_band\n",
    "    df['lower_band'] = lower_band\n",
    "    df['mid_band'] = mid_band\n",
    "    \n",
    "    df['upper_band_nine'] = upper_band_nine\n",
    "    df['lower_band_nine'] = lower_band_nine\n",
    "    df['mid_band_nine'] = mid_band_nine\n",
    "    \n",
    "    df['upper_band_one'] = upper_band_one\n",
    "    df['lower_band_one'] = lower_band_one\n",
    "    df['mid_band_one'] = mid_band_one\n",
    "    \n",
    "    df['upper_band_slow'] = upper_band_slow\n",
    "    df['lower_band_slow'] = lower_band_slow\n",
    "    df['mid_band_slow'] = mid_band_slow\n",
    "    \n",
    "    df['d_ema'] = d_ema\n",
    "    df['ht_trend'] = ht_trend\n",
    "    df['kama'] = kama\n",
    "    df['ma_fast'] = ma_fast\n",
    "    df['ma_really_fast'] = ma_really_fast\n",
    "    df['ma_slow'] = ma_slow\n",
    "    #df['mama'] = mama\n",
    "    df['mid'] = mid\n",
    "    df['mid_price'] = mid_price\n",
    "\n",
    "    df['sar'] = sar\n",
    "    df['sarext'] = sarext\n",
    "    df['sma'] = sma\n",
    "    df['tema'] = tema\n",
    "    df['trima'] = trima\n",
    "    df['wma'] = wma\n",
    "    #df['fama'] = fama\n",
    "\n",
    "    df['EMA'] = E_M_A\n",
    "    df['SlowK'] = slowk\n",
    "    df['SlowD'] = slowd\n",
    "    df['R_S_I'] = R_S_I\n",
    "    df['FastK'] = fastk\n",
    "    df['FastD'] = fastd\n",
    "    df['WilliamsR'] = real\n",
    "\n",
    "    df['atr'] = atr\n",
    "    df['natr'] = natr\n",
    "    df['t_range'] = t_range\n",
    "\n",
    "\n",
    "    #df['na_tr'] = natr\n",
    "\n",
    "    df['OBV'] = O_B_V\n",
    "    df['ADOSC'] = A_D_O_S_C\n",
    "    df['ogchaikin'] = O_G_chaikin\n",
    "    df['HTDCperiod'] = HT_DCper\n",
    "    df['HTDCphase'] = HT_DCphase\n",
    "    df['inphase'] = inphase\n",
    "    df['quad'] = quad\n",
    "    df['rsin'] = r_sin\n",
    "    df['leadsin'] = leadsin\n",
    "\n",
    "    df['mfi'] = mfi\n",
    "    df['min_di'] = min_di\n",
    "    df['min_dm'] = min_dm\n",
    "    df['momo'] = momo\n",
    "    df['plus_di'] = plus_di\n",
    "    df['plus_dm'] = plus_dm\n",
    "    df['ppo'] = ppo\n",
    "    df['roc'] = roc\n",
    "    df['rocp'] = rocp\n",
    "\n",
    "    df['rocr'] = rocr\n",
    "    df['rocr_hund'] = rocr_hund\n",
    "    df['rsi_fastk'] = rsi_fastk\n",
    "    df['rsi_fastd'] = rsi_fastd\n",
    "    df['trix'] = trix\n",
    "    df['ult_osc'] = ult_osc\n",
    "\n",
    "    df['adx'] = adx\n",
    "    df['adxr'] = adxr\n",
    "    df['apo'] = apo\n",
    "    df['aroon_d'] = aroon_d\n",
    "    df['aroon_u'] = aroon_u\n",
    "    df['aroon_osc'] = aroon_osc\n",
    "    df['bop'] = bop\n",
    "    df['cci'] = cci\n",
    "    df['cmo'] = cmo\n",
    "\n",
    "    df['dx'] = dx\n",
    "    df['macd'] = macd\n",
    "    df['macdsig'] = macdsig\n",
    "    df['macdhist'] = macdhist\n",
    "    df['macdex'] = macdex\n",
    "    df['macdexsig'] = macdexsig\n",
    "    df['macdexhist'] = macdexhist\n",
    "    df['macdfixd'] = macdfixd\n",
    "    df['macdfixdsig'] = macdfixdsig\n",
    "    df['macdfixdhist'] = macdfixdhist\n",
    "\n",
    "    df['two_crows'] = two_crows\n",
    "    df['three_crows'] = three_crows\n",
    "    df['three_inside'] = three_inside\n",
    "    df['three_line'] = three_line\n",
    "    df['three_out'] = three_out\n",
    "    df['three_stars'] = three_stars\n",
    "    df['three_soldier'] = three_soldier\n",
    "    df['baby'] = baby\n",
    "    df['adv'] = adv\n",
    "    df['belt_hold'] = belt_hold\n",
    "    df['breakaway'] = breakaway\n",
    "    df['closingmara'] = closingmara\n",
    "    df['baby_swallow'] = belt_hold\n",
    "\n",
    "    df['counter'] = counter\n",
    "    df['dark_cloud'] = dark_cloud\n",
    "    df['doji'] = doji\n",
    "    df['doji_star'] = doji_star\n",
    "    df['dragon_doji'] = dragon_doji\n",
    "    df['engulf'] = engulf\n",
    "    df['evening_star'] = evening_star\n",
    "    df['gapside'] = gapside\n",
    "    df['gravestone'] = gravestone\n",
    "    df['hammer'] = hammer\n",
    "    df['hang_man'] = hang_man\n",
    "    df['harami'] = harami\n",
    "    df['harami_cross'] = harami_cross\n",
    "\n",
    "    df['high_wave'] = high_wave\n",
    "    df['hikkake'] = hikkake\n",
    "    df['hikkake_mod'] = hikkake_mod\n",
    "    df['pidgeon'] = pidgeon\n",
    "    df['id_three_crows'] = id_three_crows\n",
    "    df['in_neck'] = in_neck\n",
    "    df['inv_hammer'] = inv_hammer\n",
    "    df['kicking'] = kicking\n",
    "    df['kicking_len'] = kicking_len\n",
    "    df['ladder_bot'] = ladder_bot\n",
    "    df['doji_long'] = doji_long\n",
    "    df['long_line'] = long_line\n",
    "    df['marabozu'] = marabozu\n",
    "                                                    # this is  a comment\n",
    "    df['match_glow'] = match_glow\n",
    "    df['mat_hold'] = mat_hold\n",
    "    df['morning_doji'] = morning_doji\n",
    "    df['morning_star'] = morning_star\n",
    "    df['on_neck'] = on_neck\n",
    "    df['pierce'] = pierce\n",
    "    df['rickshaw'] = rickshaw\n",
    "    df['rise_fall'] = rise_fall\n",
    "    df['sep_line'] = sep_line\n",
    "    df['shooting_star'] = shooting_star\n",
    "    df['sl_candle'] = sl_candle\n",
    "    df['spin_top'] = spin_top\n",
    "    df['stalled'] = stalled\n",
    "\n",
    "    df['stick_sand'] = stick_sand\n",
    "    df['takuri'] = takuri\n",
    "    df['tasuki_gap'] = tasuki_gap\n",
    "    df['thrust'] = thrust\n",
    "    df['tristar'] = tristar\n",
    "    df['three_river'] = three_river\n",
    "    df['ud_two_gap'] = ud_two_gap\n",
    "    df['down_three_gap'] = down_three_gap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "suspected-conspiracy",
   "metadata": {},
   "outputs": [],
   "source": [
    "def add_lag(num_lag_cols, this_df):\n",
    "    \n",
    "    tic = time.perf_counter()\n",
    "    \n",
    "    lag_cols = ['ema', 'slowk','slowd','r_s_i','fastk','fastd','williamsr','volume','range_hl','range_oc','adj_close',\n",
    "                'obv', 'adosc', 'ogchaikin', 'htdcperiod','htdcphase',\n",
    "                'inphase','quad','rsin','leadsin', 'two_crows', 'three_crows', 'three_inside', 'three_line',\n",
    "                'three_out', 'three_stars', 'three_soldier', 'baby', 'adv', 'belt_hold',\n",
    "                'breakaway', 'closingmara', 'baby_swallow', 'counter','dark_cloud','doji','doji_star','dragon_doji',\n",
    "                'engulf','evening_star','gapside','gravestone','hammer',\n",
    "                'hang_man','harami','harami_cross','high_wave','hikkake','hikkake_mod','pidgeon','id_three_crows',\n",
    "                'in_neck','inv_hammer','kicking','kicking_len','ladder_bot',\n",
    "                'doji_long','long_line','marabozu', 'match_glow','mat_hold','morning_doji','morning_star','on_neck',\n",
    "                'pierce','rickshaw','rise_fall','sep_line','shooting_star',\n",
    "                'sl_candle','spin_top','stalled','stick_sand','takuri','tasuki_gap','thrust','tristar','three_river',\n",
    "                'ud_two_gap','down_three_gap', 'upper_band','lower_band',\n",
    "                'mid_band','d_ema','ht_trend','kama','ma_fast','ma_really_fast','ma_slow','mid','mid_price','sar',\n",
    "                'sarext','sma','tema','trima','wma',\n",
    "                'adx','adxr','apo','aroon_d','aroon_u','aroon_osc',\n",
    "                'bop','cci','cmo','dx','macd','macdsig','macdhist','macdex','macdexsig','macdexhist','macdfixd',\n",
    "                'macdfixdsig','macdfixdhist','mfi','min_di','min_dm',\n",
    "                'momo','plus_di','plus_dm','ppo','roc','rocp','rocr','rocr_hund','rsi_fastk','rsi_fastd','trix',\n",
    "                'ult_osc', 'atr','natr','t_range',\n",
    "                'upper_band_nine','lower_band_nine','mid_band_nine','upper_band_one','lower_band_one','mid_band_one',\n",
    "                'upper_band_slow','lower_band_slow','mid_band_slow'\n",
    "                ]\n",
    "\n",
    "    shift_range = [x+1 for x in range(num_lag_cols)]\n",
    "\n",
    "    for shift in shift_range:\n",
    "        train_shift = this_df[merging_keys + lag_cols].copy()\n",
    "\n",
    "        # E.g. order_day of 0 becomes 1, for shift = 1.\n",
    "        # So when this is merged with order_day of 1 in df, this will represent lag of 1.\n",
    "        train_shift['order_day'] = train_shift['order_day'] + shift\n",
    "\n",
    "        foo = lambda x: '{}_lag_{}'.format(x, shift) if x in lag_cols else x\n",
    "        train_shift = train_shift.rename(columns=foo)\n",
    "\n",
    "        this_df = pd.merge(this_df, train_shift, on=merging_keys, how='left') #.fillna(0)\n",
    "\n",
    "    del train_shift\n",
    "    \n",
    "    toc = time.perf_counter()\n",
    "\n",
    "    tic_toc = (toc - tic) / 60\n",
    "\n",
    "    print(f\"completed lagging in {tic_toc:0.4f} min\")\n",
    "    \n",
    "\n",
    "    return this_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "impossible-steam",
   "metadata": {},
   "outputs": [],
   "source": [
    "def add_labels(df, window_size):\n",
    "    \n",
    "    tic = time.perf_counter()\n",
    "    \n",
    "    print(window_size)\n",
    "    \n",
    "    df.loc[df['adj_close'] >= df['adj_close'].rolling(window_size, center = True).quantile(.9, interpolation='linear'), 'labels'] = 0 # sell\n",
    "    \n",
    "    df.loc[df['adj_close'] <= df['adj_close'].rolling(window_size, center = True).quantile(.1, interpolation='linear'), 'labels'] = 1 #buy\n",
    "    \n",
    "    df['labels'].fillna(2, inplace = True)\n",
    "    \n",
    "    \n",
    "    toc = time.perf_counter()\n",
    "\n",
    "    tic_toc = (toc - tic) / 60\n",
    "    \n",
    "    #BUY => 1, SELL => 0, HOLD => 2\n",
    "\n",
    "    print(f\"completed labels in {tic_toc:0.4f} min\")\n",
    "    \n",
    "    #return list_test_labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "practical-reason",
   "metadata": {},
   "outputs": [],
   "source": [
    "def add_scale(num_interval_lag):\n",
    "\n",
    "    cols_to_scale = ['adj_close','ema', 'slowk','slowd','r_s_i','fastk','fastd','williamsr','volume','range_hl',\n",
    "                     'range_oc', 'obv', 'adosc', 'ogchaikin', 'htdcperiod','htdcphase',\n",
    "                    'inphase','quad','rsin','leadsin', 'two_crows', 'three_crows', 'three_inside', 'three_line',\n",
    "                     'three_out', 'three_stars', 'three_soldier', 'baby', 'adv', 'belt_hold',\n",
    "                    'breakaway', 'closingmara', 'baby_swallow', 'counter','dark_cloud','doji','doji_star',\n",
    "                     'dragon_doji','engulf','evening_star','gapside','gravestone','hammer',\n",
    "                    'hang_man','harami','harami_cross','high_wave','hikkake','hikkake_mod','pidgeon','id_three_crows',\n",
    "                     'in_neck','inv_hammer','kicking','kicking_len','ladder_bot',\n",
    "                    'doji_long','long_line','marabozu', 'match_glow','mat_hold','morning_doji','morning_star',\n",
    "                     'on_neck','pierce','rickshaw','rise_fall','sep_line','shooting_star',\n",
    "                    'sl_candle','spin_top','stalled','stick_sand','takuri','tasuki_gap','thrust','tristar',\n",
    "                     'three_river','ud_two_gap','down_three_gap', 'upper_band','lower_band',\n",
    "                    'mid_band','d_ema','ht_trend','kama','ma_fast','ma_really_fast','ma_slow','mid','mid_price',\n",
    "                     'sar','sarext','sma','tema',\n",
    "                     'trima','wma','adx','adxr','apo','aroon_d','aroon_u','aroon_osc',\n",
    "                    'bop','cci','cmo','dx','macd','macdsig','macdhist','macdex','macdexsig','macdexhist',\n",
    "                     'macdfixd','macdfixdsig','macdfixdhist','mfi','min_di','min_dm',\n",
    "                    'momo','plus_di','plus_dm','ppo','roc','rocp','rocr','rocr_hund','rsi_fastk','rsi_fastd',\n",
    "                     'trix','ult_osc', 'atr','natr','t_range',\n",
    "                    'upper_band_nine','lower_band_nine','mid_band_nine','upper_band_one','lower_band_one',\n",
    "                     'mid_band_one','upper_band_slow','lower_band_slow','mid_band_slow' \n",
    "                    ]\n",
    "\n",
    "    for i in range(1,num_interval_lag+1):\n",
    "        cols_to_scale.append(\"ema_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"slowk_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"slowd_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"r_s_i_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"fastk_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"fastd_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"williamsr_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"volume_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"range_hl_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"range_oc_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"adj_close_lag_\"+str(i))\n",
    "\n",
    "        cols_to_scale.append(\"upper_band_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"lower_band_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"mid_band_lag_\"+str(i))\n",
    "        \n",
    "        cols_to_scale.append(\"upper_band_nine_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"lower_band_nine_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"mid_band_nine_lag_\"+str(i))\n",
    "        \n",
    "        cols_to_scale.append(\"upper_band_one_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"lower_band_one_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"mid_band_one_lag_\"+str(i))\n",
    "        \n",
    "        cols_to_scale.append(\"upper_band_slow_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"lower_band_slow_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"mid_band_slow_lag_\"+str(i))\n",
    "        \n",
    "        cols_to_scale.append(\"d_ema_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"ht_trend_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"kama_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"ma_fast_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"ma_really_fast_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"ma_slow_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"mid_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"mid_price_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"sar_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"sarext_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"sma_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"tema_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"trima_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"wma_lag_\"+str(i))\n",
    "\n",
    "        cols_to_scale.append(\"atr_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"natr_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"t_range_lag_\"+str(i))\n",
    "\n",
    "        #momentum indicator lag cols\n",
    "\n",
    "        cols_to_scale.append(\"adx_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"adxr_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"apo_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"aroon_d_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"aroon_u_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"aroon_osc_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"bop_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"cci_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"cmo_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"dx_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"macd_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"macdsig_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"macdhist_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"macdex_lag_\"+str(i))\n",
    "\n",
    "        cols_to_scale.append(\"mfi_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"min_di_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"min_dm_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"momo_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"plus_di_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"plus_dm_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"ppo_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"roc_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"rocp_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"rocr_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"rocr_hund_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"rsi_fastk_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"rsi_fastd_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"trix_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"ult_osc_lag_\"+str(i))\n",
    "\n",
    "        cols_to_scale.append(\"macdexsig_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"macdexhist_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"macdfixd_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"macdfixdsig_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"macdfixdhist_lag_\"+str(i))\n",
    "\n",
    "\n",
    "        #cols_to_scale.append(\"mama_lag_\"+str(i))\n",
    "        #cols_to_scale.append(\"NATR_lag_\"+str(i))\n",
    "\n",
    "        cols_to_scale.append(\"obv_lag_\" +str(i))\n",
    "        cols_to_scale.append(\"adosc_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"ogchaikin_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"htdcperiod_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"htdcphase_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"inphase_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"quad_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"rsin_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"leadsin_lag_\"+str(i))\n",
    "        #cols_to_scale.append(\"fama_lag_\"+str(i))\n",
    "\n",
    "        cols_to_scale.append(\"two_crows_lag_\" +str(i))\n",
    "        cols_to_scale.append(\"three_crows_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"three_inside_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"three_line_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"three_out_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"three_stars_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"three_soldier_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"baby_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"adv_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"belt_hold_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"breakaway_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"closingmara_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"baby_swallow_lag_\"+str(i))\n",
    "\n",
    "        cols_to_scale.append(\"counter_lag_\" +str(i))\n",
    "        cols_to_scale.append(\"dark_cloud_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"doji_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"doji_star_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"dragon_doji_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"engulf_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"evening_star_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"gapside_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"gravestone_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"hammer_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"hang_man_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"harami_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"harami_cross_lag_\"+str(i))\n",
    "\n",
    "        cols_to_scale.append(\"high_wave_lag_\" +str(i))\n",
    "        cols_to_scale.append(\"hikkake_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"hikkake_mod_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"pidgeon_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"id_three_crows_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"in_neck_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"inv_hammer_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"kicking_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"kicking_len_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"ladder_bot_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"doji_long_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"long_line_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"marabozu_lag_\"+str(i))\n",
    "\n",
    "        cols_to_scale.append(\"match_glow_lag_\" +str(i))\n",
    "        cols_to_scale.append(\"mat_hold_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"morning_doji_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"morning_star_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"on_neck_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"pierce_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"rickshaw_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"rise_fall_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"sep_line_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"shooting_star_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"sl_candle_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"spin_top_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"stalled_lag_\"+str(i))\n",
    "\n",
    "        cols_to_scale.append(\"stick_sand_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"takuri_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"tasuki_gap_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"thrust_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"tristar_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"three_river_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"ud_two_gap_lag_\"+str(i))\n",
    "        cols_to_scale.append(\"down_three_gap_lag_\"+str(i))\n",
    "\n",
    "    return cols_to_scale"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "id": "exterior-burner",
   "metadata": {},
   "outputs": [],
   "source": [
    "stock = 'NVAX'\n",
    "df = pd.read_csv(r'C:\\Users\\Michael\\Desktop\\Python\\Stonks\\Data_Scrape\\file_storage_new\\NVAX.csv'\n",
    "                 , parse_dates = [['<DATE>', '<TIME>']])\n",
    "\n",
    "df.drop(['Unnamed: 0'], axis=1, inplace = True)\n",
    "\n",
    "\n",
    "df.columns = ['datetime','Open','High','Low','Close','Volume']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "id": "indian-sculpture",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Index(['datetime', 'Open', 'High', 'Low', 'Close', 'Volume'], dtype='object')\n"
     ]
    }
   ],
   "source": [
    "date_time = df['datetime']\n",
    "df.head(4)\n",
    "\n",
    "print(df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "id": "disciplinary-packing",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "#adj_close_list = df['Close']\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "id": "middle-maintenance",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>datetime</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2020-01-31 17:30:00</td>\n",
       "      <td>7.210</td>\n",
       "      <td>7.290</td>\n",
       "      <td>7.15</td>\n",
       "      <td>7.1896</td>\n",
       "      <td>145419</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2020-01-31 17:31:00</td>\n",
       "      <td>7.205</td>\n",
       "      <td>7.205</td>\n",
       "      <td>7.00</td>\n",
       "      <td>7.1099</td>\n",
       "      <td>57453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2020-01-31 17:32:00</td>\n",
       "      <td>7.030</td>\n",
       "      <td>7.110</td>\n",
       "      <td>7.03</td>\n",
       "      <td>7.0300</td>\n",
       "      <td>22454</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2020-01-31 17:33:00</td>\n",
       "      <td>7.030</td>\n",
       "      <td>7.060</td>\n",
       "      <td>6.93</td>\n",
       "      <td>6.9400</td>\n",
       "      <td>60747</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             datetime   Open   High   Low   Close  Volume\n",
       "0 2020-01-31 17:30:00  7.210  7.290  7.15  7.1896  145419\n",
       "1 2020-01-31 17:31:00  7.205  7.205  7.00  7.1099   57453\n",
       "2 2020-01-31 17:32:00  7.030  7.110  7.03  7.0300   22454\n",
       "3 2020-01-31 17:33:00  7.030  7.060  6.93  6.9400   60747"
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "id": "recreational-trade",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             datetime   Close  Volume\n",
      "0 2020-01-31 17:30:00  7.1896  145419\n",
      "1 2020-01-31 17:31:00  7.1099   57453\n",
      "2 2020-01-31 17:32:00  7.0300   22454\n",
      "3 2020-01-31 17:33:00  6.9400   60747\n"
     ]
    }
   ],
   "source": [
    "new_df = pd.concat([date_time, df['Close'], df['Volume']], axis = 1)\n",
    "\n",
    "print(new_df.head(4))\n",
    "\n",
    "new_df.to_csv('NVAX_test.csv', index = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "id": "cutting-vienna",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Reading data in batches:\n",
      "Batch number: 0\n",
      "Returning bars \n",
      "\n"
     ]
    }
   ],
   "source": [
    "from mlfinlab.data_structures import standard_data_structures\n",
    "\n",
    "\n",
    "# Dollar Bars\n",
    "\n",
    "dollar = standard_data_structures.get_dollar_bars('NVAX_test.csv', threshold=700,\n",
    "batch_size=1000000, verbose=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "id": "random-steps",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "            date_time  tick_num    open    high     low   close  volume  \\\n",
      "0 2020-01-31 17:30:00         1  7.1896  7.1896  7.1896  7.1896  145419   \n",
      "1 2020-01-31 17:31:00         2  7.1099  7.1099  7.1099  7.1099   57453   \n",
      "2 2020-01-31 17:32:00         3  7.0300  7.0300  7.0300  7.0300   22454   \n",
      "3 2020-01-31 17:33:00         4  6.9400  6.9400  6.9400  6.9400   60747   \n",
      "4 2020-01-31 17:34:00         5  7.0400  7.0400  7.0400  7.0400   25353   \n",
      "\n",
      "   cum_buy_volume  cum_ticks  cum_dollar_value  \n",
      "0               0          1      1.045504e+06  \n",
      "1               0          1      4.084851e+05  \n",
      "2               0          1      1.578516e+05  \n",
      "3               0          1      4.215842e+05  \n",
      "4           25353          1      1.784851e+05  \n"
     ]
    }
   ],
   "source": [
    "print(dollar.head(5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "id": "empirical-sigma",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "/* global mpl */\n",
       "window.mpl = {};\n",
       "\n",
       "mpl.get_websocket_type = function () {\n",
       "    if (typeof WebSocket !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof MozWebSocket !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert(\n",
       "            'Your browser does not have WebSocket support. ' +\n",
       "                'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "                'Firefox 4 and 5 are also supported but you ' +\n",
       "                'have to enable WebSockets in about:config.'\n",
       "        );\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure = function (figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = this.ws.binaryType !== undefined;\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById('mpl-warnings');\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent =\n",
       "                'This browser does not support binary websocket messages. ' +\n",
       "                'Performance may be slow.';\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = document.createElement('div');\n",
       "    this.root.setAttribute('style', 'display: inline-block');\n",
       "    this._root_extra_style(this.root);\n",
       "\n",
       "    parent_element.appendChild(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen = function () {\n",
       "        fig.send_message('supports_binary', { value: fig.supports_binary });\n",
       "        fig.send_message('send_image_mode', {});\n",
       "        if (fig.ratio !== 1) {\n",
       "            fig.send_message('set_dpi_ratio', { dpi_ratio: fig.ratio });\n",
       "        }\n",
       "        fig.send_message('refresh', {});\n",
       "    };\n",
       "\n",
       "    this.imageObj.onload = function () {\n",
       "        if (fig.image_mode === 'full') {\n",
       "            // Full images could contain transparency (where diff images\n",
       "            // almost always do), so we need to clear the canvas so that\n",
       "            // there is no ghosting.\n",
       "            fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "        }\n",
       "        fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "    };\n",
       "\n",
       "    this.imageObj.onunload = function () {\n",
       "        fig.ws.close();\n",
       "    };\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_header = function () {\n",
       "    var titlebar = document.createElement('div');\n",
       "    titlebar.classList =\n",
       "        'ui-dialog-titlebar ui-widget-header ui-corner-all ui-helper-clearfix';\n",
       "    var titletext = document.createElement('div');\n",
       "    titletext.classList = 'ui-dialog-title';\n",
       "    titletext.setAttribute(\n",
       "        'style',\n",
       "        'width: 100%; text-align: center; padding: 3px;'\n",
       "    );\n",
       "    titlebar.appendChild(titletext);\n",
       "    this.root.appendChild(titlebar);\n",
       "    this.header = titletext;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = (this.canvas_div = document.createElement('div'));\n",
       "    canvas_div.setAttribute(\n",
       "        'style',\n",
       "        'border: 1px solid #ddd;' +\n",
       "            'box-sizing: content-box;' +\n",
       "            'clear: both;' +\n",
       "            'min-height: 1px;' +\n",
       "            'min-width: 1px;' +\n",
       "            'outline: 0;' +\n",
       "            'overflow: hidden;' +\n",
       "            'position: relative;' +\n",
       "            'resize: both;'\n",
       "    );\n",
       "\n",
       "    function on_keyboard_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.key_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    canvas_div.addEventListener(\n",
       "        'keydown',\n",
       "        on_keyboard_event_closure('key_press')\n",
       "    );\n",
       "    canvas_div.addEventListener(\n",
       "        'keyup',\n",
       "        on_keyboard_event_closure('key_release')\n",
       "    );\n",
       "\n",
       "    this._canvas_extra_style(canvas_div);\n",
       "    this.root.appendChild(canvas_div);\n",
       "\n",
       "    var canvas = (this.canvas = document.createElement('canvas'));\n",
       "    canvas.classList.add('mpl-canvas');\n",
       "    canvas.setAttribute('style', 'box-sizing: content-box;');\n",
       "\n",
       "    this.context = canvas.getContext('2d');\n",
       "\n",
       "    var backingStore =\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        this.context.webkitBackingStorePixelRatio ||\n",
       "        this.context.mozBackingStorePixelRatio ||\n",
       "        this.context.msBackingStorePixelRatio ||\n",
       "        this.context.oBackingStorePixelRatio ||\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        1;\n",
       "\n",
       "    this.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband_canvas = (this.rubberband_canvas = document.createElement(\n",
       "        'canvas'\n",
       "    ));\n",
       "    rubberband_canvas.setAttribute(\n",
       "        'style',\n",
       "        'box-sizing: content-box; position: absolute; left: 0; top: 0; z-index: 1;'\n",
       "    );\n",
       "\n",
       "    // Apply a ponyfill if ResizeObserver is not implemented by browser.\n",
       "    if (this.ResizeObserver === undefined) {\n",
       "        if (window.ResizeObserver !== undefined) {\n",
       "            this.ResizeObserver = window.ResizeObserver;\n",
       "        } else {\n",
       "            var obs = _JSXTOOLS_RESIZE_OBSERVER({});\n",
       "            this.ResizeObserver = obs.ResizeObserver;\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.resizeObserverInstance = new this.ResizeObserver(function (entries) {\n",
       "        var nentries = entries.length;\n",
       "        for (var i = 0; i < nentries; i++) {\n",
       "            var entry = entries[i];\n",
       "            var width, height;\n",
       "            if (entry.contentBoxSize) {\n",
       "                if (entry.contentBoxSize instanceof Array) {\n",
       "                    // Chrome 84 implements new version of spec.\n",
       "                    width = entry.contentBoxSize[0].inlineSize;\n",
       "                    height = entry.contentBoxSize[0].blockSize;\n",
       "                } else {\n",
       "                    // Firefox implements old version of spec.\n",
       "                    width = entry.contentBoxSize.inlineSize;\n",
       "                    height = entry.contentBoxSize.blockSize;\n",
       "                }\n",
       "            } else {\n",
       "                // Chrome <84 implements even older version of spec.\n",
       "                width = entry.contentRect.width;\n",
       "                height = entry.contentRect.height;\n",
       "            }\n",
       "\n",
       "            // Keep the size of the canvas and rubber band canvas in sync with\n",
       "            // the canvas container.\n",
       "            if (entry.devicePixelContentBoxSize) {\n",
       "                // Chrome 84 implements new version of spec.\n",
       "                canvas.setAttribute(\n",
       "                    'width',\n",
       "                    entry.devicePixelContentBoxSize[0].inlineSize\n",
       "                );\n",
       "                canvas.setAttribute(\n",
       "                    'height',\n",
       "                    entry.devicePixelContentBoxSize[0].blockSize\n",
       "                );\n",
       "            } else {\n",
       "                canvas.setAttribute('width', width * fig.ratio);\n",
       "                canvas.setAttribute('height', height * fig.ratio);\n",
       "            }\n",
       "            canvas.setAttribute(\n",
       "                'style',\n",
       "                'width: ' + width + 'px; height: ' + height + 'px;'\n",
       "            );\n",
       "\n",
       "            rubberband_canvas.setAttribute('width', width);\n",
       "            rubberband_canvas.setAttribute('height', height);\n",
       "\n",
       "            // And update the size in Python. We ignore the initial 0/0 size\n",
       "            // that occurs as the element is placed into the DOM, which should\n",
       "            // otherwise not happen due to the minimum size styling.\n",
       "            if (fig.ws.readyState == 1 && width != 0 && height != 0) {\n",
       "                fig.request_resize(width, height);\n",
       "            }\n",
       "        }\n",
       "    });\n",
       "    this.resizeObserverInstance.observe(canvas_div);\n",
       "\n",
       "    function on_mouse_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.mouse_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousedown',\n",
       "        on_mouse_event_closure('button_press')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseup',\n",
       "        on_mouse_event_closure('button_release')\n",
       "    );\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousemove',\n",
       "        on_mouse_event_closure('motion_notify')\n",
       "    );\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseenter',\n",
       "        on_mouse_event_closure('figure_enter')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseleave',\n",
       "        on_mouse_event_closure('figure_leave')\n",
       "    );\n",
       "\n",
       "    canvas_div.addEventListener('wheel', function (event) {\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        on_mouse_event_closure('scroll')(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.appendChild(canvas);\n",
       "    canvas_div.appendChild(rubberband_canvas);\n",
       "\n",
       "    this.rubberband_context = rubberband_canvas.getContext('2d');\n",
       "    this.rubberband_context.strokeStyle = '#000000';\n",
       "\n",
       "    this._resize_canvas = function (width, height, forward) {\n",
       "        if (forward) {\n",
       "            canvas_div.style.width = width + 'px';\n",
       "            canvas_div.style.height = height + 'px';\n",
       "        }\n",
       "    };\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    this.rubberband_canvas.addEventListener('contextmenu', function (_e) {\n",
       "        event.preventDefault();\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus() {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'mpl-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'mpl-button-group';\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'mpl-button-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        var button = (fig.buttons[name] = document.createElement('button'));\n",
       "        button.classList = 'mpl-widget';\n",
       "        button.setAttribute('role', 'button');\n",
       "        button.setAttribute('aria-disabled', 'false');\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "\n",
       "        var icon_img = document.createElement('img');\n",
       "        icon_img.src = '_images/' + image + '.png';\n",
       "        icon_img.srcset = '_images/' + image + '_large.png 2x';\n",
       "        icon_img.alt = tooltip;\n",
       "        button.appendChild(icon_img);\n",
       "\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    var fmt_picker = document.createElement('select');\n",
       "    fmt_picker.classList = 'mpl-widget';\n",
       "    toolbar.appendChild(fmt_picker);\n",
       "    this.format_dropdown = fmt_picker;\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = document.createElement('option');\n",
       "        option.selected = fmt === mpl.default_extension;\n",
       "        option.innerHTML = fmt;\n",
       "        fmt_picker.appendChild(option);\n",
       "    }\n",
       "\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.request_resize = function (x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', { width: x_pixels, height: y_pixels });\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_message = function (type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function () {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({ type: 'draw', figure_id: this.id }));\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function (fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] !== fig.canvas.width || size[1] !== fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1], msg['forward']);\n",
       "        fig.send_message('refresh', {});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function (fig, msg) {\n",
       "    var x0 = msg['x0'] / fig.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / fig.ratio;\n",
       "    var x1 = msg['x1'] / fig.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / fig.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0,\n",
       "        0,\n",
       "        fig.canvas.width / fig.ratio,\n",
       "        fig.canvas.height / fig.ratio\n",
       "    );\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function (fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function (fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch (cursor) {\n",
       "        case 0:\n",
       "            cursor = 'pointer';\n",
       "            break;\n",
       "        case 1:\n",
       "            cursor = 'default';\n",
       "            break;\n",
       "        case 2:\n",
       "            cursor = 'crosshair';\n",
       "            break;\n",
       "        case 3:\n",
       "            cursor = 'move';\n",
       "            break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_message = function (fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function (fig, _msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function (fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_history_buttons = function (fig, msg) {\n",
       "    for (var key in msg) {\n",
       "        if (!(key in fig.buttons)) {\n",
       "            continue;\n",
       "        }\n",
       "        fig.buttons[key].disabled = !msg[key];\n",
       "        fig.buttons[key].setAttribute('aria-disabled', !msg[key]);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_navigate_mode = function (fig, msg) {\n",
       "    if (msg['mode'] === 'PAN') {\n",
       "        fig.buttons['Pan'].classList.add('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    } else if (msg['mode'] === 'ZOOM') {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.add('active');\n",
       "    } else {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message('ack', {});\n",
       "};\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function (fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = 'image/png';\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src\n",
       "                );\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data\n",
       "            );\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        } else if (\n",
       "            typeof evt.data === 'string' &&\n",
       "            evt.data.slice(0, 21) === 'data:image/png;base64'\n",
       "        ) {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig['handle_' + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\n",
       "                \"No handler for the '\" + msg_type + \"' message type: \",\n",
       "                msg\n",
       "            );\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\n",
       "                    \"Exception inside the 'handler_\" + msg_type + \"' callback:\",\n",
       "                    e,\n",
       "                    e.stack,\n",
       "                    msg\n",
       "                );\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "};\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function (e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e) {\n",
       "        e = window.event;\n",
       "    }\n",
       "    if (e.target) {\n",
       "        targ = e.target;\n",
       "    } else if (e.srcElement) {\n",
       "        targ = e.srcElement;\n",
       "    }\n",
       "    if (targ.nodeType === 3) {\n",
       "        // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "    }\n",
       "\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    var boundingRect = targ.getBoundingClientRect();\n",
       "    var x = e.pageX - (boundingRect.left + document.body.scrollLeft);\n",
       "    var y = e.pageY - (boundingRect.top + document.body.scrollTop);\n",
       "\n",
       "    return { x: x, y: y };\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys(original) {\n",
       "    return Object.keys(original).reduce(function (obj, key) {\n",
       "        if (typeof original[key] !== 'object') {\n",
       "            obj[key] = original[key];\n",
       "        }\n",
       "        return obj;\n",
       "    }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function (event, name) {\n",
       "    var canvas_pos = mpl.findpos(event);\n",
       "\n",
       "    if (name === 'button_press') {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * this.ratio;\n",
       "    var y = canvas_pos.y * this.ratio;\n",
       "\n",
       "    this.send_message(name, {\n",
       "        x: x,\n",
       "        y: y,\n",
       "        button: event.button,\n",
       "        step: event.step,\n",
       "        guiEvent: simpleKeys(event),\n",
       "    });\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (_event, _name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.key_event = function (event, name) {\n",
       "    // Prevent repeat events\n",
       "    if (name === 'key_press') {\n",
       "        if (event.which === this._key) {\n",
       "            return;\n",
       "        } else {\n",
       "            this._key = event.which;\n",
       "        }\n",
       "    }\n",
       "    if (name === 'key_release') {\n",
       "        this._key = null;\n",
       "    }\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which !== 17) {\n",
       "        value += 'ctrl+';\n",
       "    }\n",
       "    if (event.altKey && event.which !== 18) {\n",
       "        value += 'alt+';\n",
       "    }\n",
       "    if (event.shiftKey && event.which !== 16) {\n",
       "        value += 'shift+';\n",
       "    }\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, { key: value, guiEvent: simpleKeys(event) });\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function (name) {\n",
       "    if (name === 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message('toolbar_button', { name: name });\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function (tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "\n",
       "///////////////// REMAINING CONTENT GENERATED BY embed_js.py /////////////////\n",
       "// prettier-ignore\n",
       "var _JSXTOOLS_RESIZE_OBSERVER=function(A){var t,i=new WeakMap,n=new WeakMap,a=new WeakMap,r=new WeakMap,o=new Set;function s(e){if(!(this instanceof s))throw new TypeError(\"Constructor requires 'new' operator\");i.set(this,e)}function h(){throw new TypeError(\"Function is not a constructor\")}function c(e,t,i,n){e=0 in arguments?Number(arguments[0]):0,t=1 in arguments?Number(arguments[1]):0,i=2 in arguments?Number(arguments[2]):0,n=3 in arguments?Number(arguments[3]):0,this.right=(this.x=this.left=e)+(this.width=i),this.bottom=(this.y=this.top=t)+(this.height=n),Object.freeze(this)}function d(){t=requestAnimationFrame(d);var s=new WeakMap,p=new Set;o.forEach((function(t){r.get(t).forEach((function(i){var r=t instanceof window.SVGElement,o=a.get(t),d=r?0:parseFloat(o.paddingTop),f=r?0:parseFloat(o.paddingRight),l=r?0:parseFloat(o.paddingBottom),u=r?0:parseFloat(o.paddingLeft),g=r?0:parseFloat(o.borderTopWidth),m=r?0:parseFloat(o.borderRightWidth),w=r?0:parseFloat(o.borderBottomWidth),b=u+f,F=d+l,v=(r?0:parseFloat(o.borderLeftWidth))+m,W=g+w,y=r?0:t.offsetHeight-W-t.clientHeight,E=r?0:t.offsetWidth-v-t.clientWidth,R=b+v,z=F+W,M=r?t.width:parseFloat(o.width)-R-E,O=r?t.height:parseFloat(o.height)-z-y;if(n.has(t)){var k=n.get(t);if(k[0]===M&&k[1]===O)return}n.set(t,[M,O]);var S=Object.create(h.prototype);S.target=t,S.contentRect=new c(u,d,M,O),s.has(i)||(s.set(i,[]),p.add(i)),s.get(i).push(S)}))})),p.forEach((function(e){i.get(e).call(e,s.get(e),e)}))}return s.prototype.observe=function(i){if(i instanceof window.Element){r.has(i)||(r.set(i,new Set),o.add(i),a.set(i,window.getComputedStyle(i)));var n=r.get(i);n.has(this)||n.add(this),cancelAnimationFrame(t),t=requestAnimationFrame(d)}},s.prototype.unobserve=function(i){if(i instanceof window.Element&&r.has(i)){var n=r.get(i);n.has(this)&&(n.delete(this),n.size||(r.delete(i),o.delete(i))),n.size||r.delete(i),o.size||cancelAnimationFrame(t)}},A.DOMRectReadOnly=c,A.ResizeObserver=s,A.ResizeObserverEntry=h,A}; // eslint-disable-line\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Left button pans, Right button zooms\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";/* global mpl */\n",
       "\n",
       "var comm_websocket_adapter = function (comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function () {\n",
       "        comm.close();\n",
       "    };\n",
       "    ws.send = function (m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function (msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data']);\n",
       "    });\n",
       "    return ws;\n",
       "};\n",
       "\n",
       "mpl.mpl_figure_comm = function (comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = document.getElementById(id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm);\n",
       "\n",
       "    function ondownload(figure, _format) {\n",
       "        window.open(figure.canvas.toDataURL());\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy, ondownload, element);\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element;\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error('Failed to find cell for figure', id, fig);\n",
       "        return;\n",
       "    }\n",
       "    fig.cell_info[0].output_area.element.on(\n",
       "        'cleared',\n",
       "        { fig: fig },\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function (fig, msg) {\n",
       "    var width = fig.canvas.width / fig.ratio;\n",
       "    fig.cell_info[0].output_area.element.off(\n",
       "        'cleared',\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "    fig.resizeObserverInstance.unobserve(fig.canvas_div);\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable();\n",
       "    fig.parent_element.innerHTML =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "    fig.close_ws(fig, msg);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.close_ws = function (fig, msg) {\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function (_remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width / this.ratio;\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message('ack', {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () {\n",
       "        fig.push_to_output();\n",
       "    }, 1000);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'btn-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'btn-group';\n",
       "    var button;\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'btn-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        button = fig.buttons[name] = document.createElement('button');\n",
       "        button.classList = 'btn btn-default';\n",
       "        button.href = '#';\n",
       "        button.title = name;\n",
       "        button.innerHTML = '<i class=\"fa ' + image + ' fa-lg\"></i>';\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message pull-right';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = document.createElement('div');\n",
       "    buttongrp.classList = 'btn-group inline pull-right';\n",
       "    button = document.createElement('button');\n",
       "    button.classList = 'btn btn-mini btn-primary';\n",
       "    button.href = '#';\n",
       "    button.title = 'Stop Interaction';\n",
       "    button.innerHTML = '<i class=\"fa fa-power-off icon-remove icon-large\"></i>';\n",
       "    button.addEventListener('click', function (_evt) {\n",
       "        fig.handle_close(fig, {});\n",
       "    });\n",
       "    button.addEventListener(\n",
       "        'mouseover',\n",
       "        on_mouseover_closure('Stop Interaction')\n",
       "    );\n",
       "    buttongrp.appendChild(button);\n",
       "    var titlebar = this.root.querySelector('.ui-dialog-titlebar');\n",
       "    titlebar.insertBefore(buttongrp, titlebar.firstChild);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._remove_fig_handler = function (event) {\n",
       "    var fig = event.data.fig;\n",
       "    if (event.target !== this) {\n",
       "        // Ignore bubbled events from children.\n",
       "        return;\n",
       "    }\n",
       "    fig.close_ws(fig, {});\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (el) {\n",
       "    el.style.boxSizing = 'content-box'; // override notebook setting of border-box.\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (el) {\n",
       "    // this is important to make the div 'focusable\n",
       "    el.setAttribute('tabindex', 0);\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    } else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (event, _name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager) {\n",
       "        manager = IPython.keyboard_manager;\n",
       "    }\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which === 13) {\n",
       "        this.canvas_div.blur();\n",
       "        // select the cell after this one\n",
       "        var index = IPython.notebook.find_cell_index(this.cell_info[0]);\n",
       "        IPython.notebook.select(index + 1);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "};\n",
       "\n",
       "mpl.find_output_cell = function (html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i = 0; i < ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code') {\n",
       "            for (var j = 0; j < cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] === html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "};\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel !== null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target(\n",
       "        'matplotlib',\n",
       "        mpl.mpl_figure_comm\n",
       "    );\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"640\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1808cbc7d60>]"
      ]
     },
     "execution_count": 146,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%matplotlib notebook\n",
    "plt.plot(dollar['tick_num'], dollar['close'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "sapphire-screening",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "id": "atlantic-benefit",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Reading data in batches:\n",
      "Batch number: 0\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\program files\\python38\\lib\\site-packages\\mlfinlab-0.10.0-py3.8.egg\\mlfinlab\\data_structures\\base_bars.py:387: RuntimeWarning: invalid value encountered in greater\n",
      "  if np.abs(self.thresholds['cum_theta']) > self.thresholds['exp_num_ticks'] * np.abs(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Returning bars \n",
      "\n"
     ]
    }
   ],
   "source": [
    "from mlfinlab.data_structures import imbalance_data_structures\n",
    "\n",
    "\n",
    "# Dollar Bars\n",
    "\n",
    "#dollar = standard_data_structures.get_dollar_bars('NVAX_test.csv', threshold=700,\n",
    "#batch_size=1000000, verbose=True)\n",
    "\n",
    "[one, volume_imbalance] = imbalance_data_structures.get_const_volume_imbalance_bars('NVAX_test.csv',\n",
    "exp_num_ticks_init=100000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "id": "crazy-latvia",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: []\n",
       "Index: []"
      ]
     },
     "execution_count": 148,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "volume_imbalance.head(4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "id": "earned-marriage",
   "metadata": {},
   "outputs": [
    {
     "ename": "KeyError",
     "evalue": "'tick_num'",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyError\u001b[0m                                  Traceback (most recent call last)",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python38\\site-packages\\pandas\\core\\indexes\\base.py\u001b[0m in \u001b[0;36mget_loc\u001b[1;34m(self, key, method, tolerance)\u001b[0m\n\u001b[0;32m   3079\u001b[0m             \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 3080\u001b[1;33m                 \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_engine\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_loc\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mcasted_key\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   3081\u001b[0m             \u001b[1;32mexcept\u001b[0m \u001b[0mKeyError\u001b[0m \u001b[1;32mas\u001b[0m \u001b[0merr\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mpandas\\_libs\\index.pyx\u001b[0m in \u001b[0;36mpandas._libs.index.IndexEngine.get_loc\u001b[1;34m()\u001b[0m\n",
      "\u001b[1;32mpandas\\_libs\\index.pyx\u001b[0m in \u001b[0;36mpandas._libs.index.IndexEngine.get_loc\u001b[1;34m()\u001b[0m\n",
      "\u001b[1;32mpandas\\_libs\\hashtable_class_helper.pxi\u001b[0m in \u001b[0;36mpandas._libs.hashtable.PyObjectHashTable.get_item\u001b[1;34m()\u001b[0m\n",
      "\u001b[1;32mpandas\\_libs\\hashtable_class_helper.pxi\u001b[0m in \u001b[0;36mpandas._libs.hashtable.PyObjectHashTable.get_item\u001b[1;34m()\u001b[0m\n",
      "\u001b[1;31mKeyError\u001b[0m: 'tick_num'",
      "\nThe above exception was the direct cause of the following exception:\n",
      "\u001b[1;31mKeyError\u001b[0m                                  Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-149-29927eed5186>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m      1\u001b[0m \u001b[0mget_ipython\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mrun_line_magic\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'matplotlib'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'notebook'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 2\u001b[1;33m \u001b[0mplt\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mvolume_imbalance\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'tick_num'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mvolume_imbalance\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'close'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python38\\site-packages\\pandas\\core\\frame.py\u001b[0m in \u001b[0;36m__getitem__\u001b[1;34m(self, key)\u001b[0m\n\u001b[0;32m   3022\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcolumns\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mnlevels\u001b[0m \u001b[1;33m>\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   3023\u001b[0m                 \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_getitem_multilevel\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 3024\u001b[1;33m             \u001b[0mindexer\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcolumns\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_loc\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   3025\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0mis_integer\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mindexer\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   3026\u001b[0m                 \u001b[0mindexer\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0mindexer\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python38\\site-packages\\pandas\\core\\indexes\\base.py\u001b[0m in \u001b[0;36mget_loc\u001b[1;34m(self, key, method, tolerance)\u001b[0m\n\u001b[0;32m   3080\u001b[0m                 \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_engine\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_loc\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mcasted_key\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   3081\u001b[0m             \u001b[1;32mexcept\u001b[0m \u001b[0mKeyError\u001b[0m \u001b[1;32mas\u001b[0m \u001b[0merr\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 3082\u001b[1;33m                 \u001b[1;32mraise\u001b[0m \u001b[0mKeyError\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mfrom\u001b[0m \u001b[0merr\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   3083\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   3084\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mtolerance\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mKeyError\u001b[0m: 'tick_num'"
     ]
    }
   ],
   "source": [
    "%matplotlib notebook\n",
    "plt.plot(volume_imbalance['tick_num'], volume_imbalance['close'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "id": "connected-herald",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    7.1896\n",
       "1    7.1099\n",
       "2    7.0300\n",
       "3    6.9400\n",
       "4    7.0400\n",
       "Name: Close, dtype: float64"
      ]
     },
     "execution_count": 150,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['Close'].head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "id": "binary-providence",
   "metadata": {},
   "outputs": [],
   "source": [
    "from mlfinlab.filters import cusum_filter\n",
    "cusum_events = cusum_filter(df['Close'], threshold=0.05, time_stamps = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "id": "unlike-kitty",
   "metadata": {},
   "outputs": [],
   "source": [
    "#print(cusum_events)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "id": "personal-aquarium",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "               datetime  Open    High     Low   Close  Volume\n",
      "81  2020-01-31 18:51:00  6.86  6.8600  6.8199  6.8300   18110\n",
      "221 2020-01-31 21:12:00  7.14  7.1573  7.1353  7.1555    8744\n",
      "292 2020-01-31 22:23:00  7.45  7.4900  7.4450  7.4800   35301\n",
      "389 2020-02-03 17:30:00  7.26  7.3000  7.2000  7.2500  215286\n",
      "509 2020-02-03 19:30:00  7.05  7.0501  7.0000  7.0200   78677\n"
     ]
    }
   ],
   "source": [
    "events_df = df.filter(items = cusum_events, axis = 0)\n",
    "\n",
    "print(events_df.head(5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "nearby-trance",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "id": "photographic-uruguay",
   "metadata": {},
   "outputs": [],
   "source": [
    "from mlfinlab.filters import z_score_filter\n",
    "z_score_events = z_score_filter(df['Close'], mean_window=100, std_window=100, z_score=2.5,\n",
    "                                time_stamps = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "id": "lined-music",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "               datetime    Open    High     Low  Close  Volume\n",
      "202 2020-01-31 20:53:00  6.9700  6.9800  6.9662  6.980    4898\n",
      "203 2020-01-31 20:54:00  6.9704  7.0500  6.9704  7.030   47852\n",
      "204 2020-01-31 20:55:00  7.0350  7.0784  7.0350  7.070   13693\n",
      "205 2020-01-31 20:56:00  7.0700  7.0708  7.0484  7.068   12740\n",
      "206 2020-01-31 20:57:00  7.0400  7.0900  7.0400  7.090    2882\n"
     ]
    }
   ],
   "source": [
    "events_df = df.filter(items = z_score_events, axis = 0)\n",
    "\n",
    "print(events_df.head(5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "id": "owned-western",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.set_index(['datetime'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "id": "changed-carnival",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "datetime\n",
      "2020-01-31 17:30:00    7.1896\n",
      "2020-01-31 17:31:00    7.1099\n",
      "2020-01-31 17:32:00    7.0300\n",
      "2020-01-31 17:33:00    6.9400\n",
      "Name: Close, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "print(df['Close'].head(4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "id": "considered-karen",
   "metadata": {},
   "outputs": [],
   "source": [
    "import mlfinlab as ml\n",
    "\n",
    "# Compute daily volatility\n",
    "\n",
    "daily_vol = ml.util.get_daily_vol(df['Close'], lookback=50)\n",
    "\n",
    "# Apply Symmetric CUSUM Filter and get timestamps for events\n",
    "# Note: Only the CUSUM filter needs a point estimate for volatility\n",
    "\n",
    "cusum_events = ml.filters.cusum_filter(df['Close'],\n",
    "                                       threshold=daily_vol[:].mean()*0.5)\n",
    "\n",
    "# Compute vertical barrier\n",
    "\n",
    "vertical_barriers = ml.labeling.add_vertical_barrier(t_events=cusum_events,\n",
    "close=df['Close'], num_days=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "id": "complete-portfolio",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2021-12-23 15:20:35.735004 33.33% apply_pt_sl_on_t1 done after 0.25 minutes. Remaining 0.49 minutes.\r",
      "2021-12-23 15:20:35.773630 66.67% apply_pt_sl_on_t1 done after 0.25 minutes. Remaining 0.12 minutes.\r",
      "2021-12-23 15:20:35.816206 100.0% apply_pt_sl_on_t1 done after 0.25 minutes. Remaining 0.0 minutes.\n"
     ]
    }
   ],
   "source": [
    "pt_sl = [1, 2]\n",
    "min_ret = 0.005\n",
    "triple_barrier_events = ml.labeling.get_events(df['Close'],\n",
    "t_events=cusum_events,\n",
    "pt_sl=pt_sl,\n",
    "target=daily_vol,\n",
    "min_ret=min_ret,\n",
    "num_threads=3,\n",
    "vertical_barrier_times=vertical_barriers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "martial-croatia",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "/* global mpl */\n",
       "window.mpl = {};\n",
       "\n",
       "mpl.get_websocket_type = function () {\n",
       "    if (typeof WebSocket !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof MozWebSocket !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert(\n",
       "            'Your browser does not have WebSocket support. ' +\n",
       "                'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "                'Firefox 4 and 5 are also supported but you ' +\n",
       "                'have to enable WebSockets in about:config.'\n",
       "        );\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure = function (figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = this.ws.binaryType !== undefined;\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById('mpl-warnings');\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent =\n",
       "                'This browser does not support binary websocket messages. ' +\n",
       "                'Performance may be slow.';\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = document.createElement('div');\n",
       "    this.root.setAttribute('style', 'display: inline-block');\n",
       "    this._root_extra_style(this.root);\n",
       "\n",
       "    parent_element.appendChild(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen = function () {\n",
       "        fig.send_message('supports_binary', { value: fig.supports_binary });\n",
       "        fig.send_message('send_image_mode', {});\n",
       "        if (fig.ratio !== 1) {\n",
       "            fig.send_message('set_dpi_ratio', { dpi_ratio: fig.ratio });\n",
       "        }\n",
       "        fig.send_message('refresh', {});\n",
       "    };\n",
       "\n",
       "    this.imageObj.onload = function () {\n",
       "        if (fig.image_mode === 'full') {\n",
       "            // Full images could contain transparency (where diff images\n",
       "            // almost always do), so we need to clear the canvas so that\n",
       "            // there is no ghosting.\n",
       "            fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "        }\n",
       "        fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "    };\n",
       "\n",
       "    this.imageObj.onunload = function () {\n",
       "        fig.ws.close();\n",
       "    };\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_header = function () {\n",
       "    var titlebar = document.createElement('div');\n",
       "    titlebar.classList =\n",
       "        'ui-dialog-titlebar ui-widget-header ui-corner-all ui-helper-clearfix';\n",
       "    var titletext = document.createElement('div');\n",
       "    titletext.classList = 'ui-dialog-title';\n",
       "    titletext.setAttribute(\n",
       "        'style',\n",
       "        'width: 100%; text-align: center; padding: 3px;'\n",
       "    );\n",
       "    titlebar.appendChild(titletext);\n",
       "    this.root.appendChild(titlebar);\n",
       "    this.header = titletext;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = (this.canvas_div = document.createElement('div'));\n",
       "    canvas_div.setAttribute(\n",
       "        'style',\n",
       "        'border: 1px solid #ddd;' +\n",
       "            'box-sizing: content-box;' +\n",
       "            'clear: both;' +\n",
       "            'min-height: 1px;' +\n",
       "            'min-width: 1px;' +\n",
       "            'outline: 0;' +\n",
       "            'overflow: hidden;' +\n",
       "            'position: relative;' +\n",
       "            'resize: both;'\n",
       "    );\n",
       "\n",
       "    function on_keyboard_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.key_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    canvas_div.addEventListener(\n",
       "        'keydown',\n",
       "        on_keyboard_event_closure('key_press')\n",
       "    );\n",
       "    canvas_div.addEventListener(\n",
       "        'keyup',\n",
       "        on_keyboard_event_closure('key_release')\n",
       "    );\n",
       "\n",
       "    this._canvas_extra_style(canvas_div);\n",
       "    this.root.appendChild(canvas_div);\n",
       "\n",
       "    var canvas = (this.canvas = document.createElement('canvas'));\n",
       "    canvas.classList.add('mpl-canvas');\n",
       "    canvas.setAttribute('style', 'box-sizing: content-box;');\n",
       "\n",
       "    this.context = canvas.getContext('2d');\n",
       "\n",
       "    var backingStore =\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        this.context.webkitBackingStorePixelRatio ||\n",
       "        this.context.mozBackingStorePixelRatio ||\n",
       "        this.context.msBackingStorePixelRatio ||\n",
       "        this.context.oBackingStorePixelRatio ||\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        1;\n",
       "\n",
       "    this.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband_canvas = (this.rubberband_canvas = document.createElement(\n",
       "        'canvas'\n",
       "    ));\n",
       "    rubberband_canvas.setAttribute(\n",
       "        'style',\n",
       "        'box-sizing: content-box; position: absolute; left: 0; top: 0; z-index: 1;'\n",
       "    );\n",
       "\n",
       "    // Apply a ponyfill if ResizeObserver is not implemented by browser.\n",
       "    if (this.ResizeObserver === undefined) {\n",
       "        if (window.ResizeObserver !== undefined) {\n",
       "            this.ResizeObserver = window.ResizeObserver;\n",
       "        } else {\n",
       "            var obs = _JSXTOOLS_RESIZE_OBSERVER({});\n",
       "            this.ResizeObserver = obs.ResizeObserver;\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.resizeObserverInstance = new this.ResizeObserver(function (entries) {\n",
       "        var nentries = entries.length;\n",
       "        for (var i = 0; i < nentries; i++) {\n",
       "            var entry = entries[i];\n",
       "            var width, height;\n",
       "            if (entry.contentBoxSize) {\n",
       "                if (entry.contentBoxSize instanceof Array) {\n",
       "                    // Chrome 84 implements new version of spec.\n",
       "                    width = entry.contentBoxSize[0].inlineSize;\n",
       "                    height = entry.contentBoxSize[0].blockSize;\n",
       "                } else {\n",
       "                    // Firefox implements old version of spec.\n",
       "                    width = entry.contentBoxSize.inlineSize;\n",
       "                    height = entry.contentBoxSize.blockSize;\n",
       "                }\n",
       "            } else {\n",
       "                // Chrome <84 implements even older version of spec.\n",
       "                width = entry.contentRect.width;\n",
       "                height = entry.contentRect.height;\n",
       "            }\n",
       "\n",
       "            // Keep the size of the canvas and rubber band canvas in sync with\n",
       "            // the canvas container.\n",
       "            if (entry.devicePixelContentBoxSize) {\n",
       "                // Chrome 84 implements new version of spec.\n",
       "                canvas.setAttribute(\n",
       "                    'width',\n",
       "                    entry.devicePixelContentBoxSize[0].inlineSize\n",
       "                );\n",
       "                canvas.setAttribute(\n",
       "                    'height',\n",
       "                    entry.devicePixelContentBoxSize[0].blockSize\n",
       "                );\n",
       "            } else {\n",
       "                canvas.setAttribute('width', width * fig.ratio);\n",
       "                canvas.setAttribute('height', height * fig.ratio);\n",
       "            }\n",
       "            canvas.setAttribute(\n",
       "                'style',\n",
       "                'width: ' + width + 'px; height: ' + height + 'px;'\n",
       "            );\n",
       "\n",
       "            rubberband_canvas.setAttribute('width', width);\n",
       "            rubberband_canvas.setAttribute('height', height);\n",
       "\n",
       "            // And update the size in Python. We ignore the initial 0/0 size\n",
       "            // that occurs as the element is placed into the DOM, which should\n",
       "            // otherwise not happen due to the minimum size styling.\n",
       "            if (fig.ws.readyState == 1 && width != 0 && height != 0) {\n",
       "                fig.request_resize(width, height);\n",
       "            }\n",
       "        }\n",
       "    });\n",
       "    this.resizeObserverInstance.observe(canvas_div);\n",
       "\n",
       "    function on_mouse_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.mouse_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousedown',\n",
       "        on_mouse_event_closure('button_press')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseup',\n",
       "        on_mouse_event_closure('button_release')\n",
       "    );\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousemove',\n",
       "        on_mouse_event_closure('motion_notify')\n",
       "    );\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseenter',\n",
       "        on_mouse_event_closure('figure_enter')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseleave',\n",
       "        on_mouse_event_closure('figure_leave')\n",
       "    );\n",
       "\n",
       "    canvas_div.addEventListener('wheel', function (event) {\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        on_mouse_event_closure('scroll')(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.appendChild(canvas);\n",
       "    canvas_div.appendChild(rubberband_canvas);\n",
       "\n",
       "    this.rubberband_context = rubberband_canvas.getContext('2d');\n",
       "    this.rubberband_context.strokeStyle = '#000000';\n",
       "\n",
       "    this._resize_canvas = function (width, height, forward) {\n",
       "        if (forward) {\n",
       "            canvas_div.style.width = width + 'px';\n",
       "            canvas_div.style.height = height + 'px';\n",
       "        }\n",
       "    };\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    this.rubberband_canvas.addEventListener('contextmenu', function (_e) {\n",
       "        event.preventDefault();\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus() {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'mpl-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'mpl-button-group';\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'mpl-button-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        var button = (fig.buttons[name] = document.createElement('button'));\n",
       "        button.classList = 'mpl-widget';\n",
       "        button.setAttribute('role', 'button');\n",
       "        button.setAttribute('aria-disabled', 'false');\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "\n",
       "        var icon_img = document.createElement('img');\n",
       "        icon_img.src = '_images/' + image + '.png';\n",
       "        icon_img.srcset = '_images/' + image + '_large.png 2x';\n",
       "        icon_img.alt = tooltip;\n",
       "        button.appendChild(icon_img);\n",
       "\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    var fmt_picker = document.createElement('select');\n",
       "    fmt_picker.classList = 'mpl-widget';\n",
       "    toolbar.appendChild(fmt_picker);\n",
       "    this.format_dropdown = fmt_picker;\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = document.createElement('option');\n",
       "        option.selected = fmt === mpl.default_extension;\n",
       "        option.innerHTML = fmt;\n",
       "        fmt_picker.appendChild(option);\n",
       "    }\n",
       "\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.request_resize = function (x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', { width: x_pixels, height: y_pixels });\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_message = function (type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function () {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({ type: 'draw', figure_id: this.id }));\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function (fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] !== fig.canvas.width || size[1] !== fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1], msg['forward']);\n",
       "        fig.send_message('refresh', {});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function (fig, msg) {\n",
       "    var x0 = msg['x0'] / fig.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / fig.ratio;\n",
       "    var x1 = msg['x1'] / fig.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / fig.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0,\n",
       "        0,\n",
       "        fig.canvas.width / fig.ratio,\n",
       "        fig.canvas.height / fig.ratio\n",
       "    );\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function (fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function (fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch (cursor) {\n",
       "        case 0:\n",
       "            cursor = 'pointer';\n",
       "            break;\n",
       "        case 1:\n",
       "            cursor = 'default';\n",
       "            break;\n",
       "        case 2:\n",
       "            cursor = 'crosshair';\n",
       "            break;\n",
       "        case 3:\n",
       "            cursor = 'move';\n",
       "            break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_message = function (fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function (fig, _msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function (fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_history_buttons = function (fig, msg) {\n",
       "    for (var key in msg) {\n",
       "        if (!(key in fig.buttons)) {\n",
       "            continue;\n",
       "        }\n",
       "        fig.buttons[key].disabled = !msg[key];\n",
       "        fig.buttons[key].setAttribute('aria-disabled', !msg[key]);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_navigate_mode = function (fig, msg) {\n",
       "    if (msg['mode'] === 'PAN') {\n",
       "        fig.buttons['Pan'].classList.add('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    } else if (msg['mode'] === 'ZOOM') {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.add('active');\n",
       "    } else {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message('ack', {});\n",
       "};\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function (fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = 'image/png';\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src\n",
       "                );\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data\n",
       "            );\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        } else if (\n",
       "            typeof evt.data === 'string' &&\n",
       "            evt.data.slice(0, 21) === 'data:image/png;base64'\n",
       "        ) {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig['handle_' + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\n",
       "                \"No handler for the '\" + msg_type + \"' message type: \",\n",
       "                msg\n",
       "            );\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\n",
       "                    \"Exception inside the 'handler_\" + msg_type + \"' callback:\",\n",
       "                    e,\n",
       "                    e.stack,\n",
       "                    msg\n",
       "                );\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "};\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function (e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e) {\n",
       "        e = window.event;\n",
       "    }\n",
       "    if (e.target) {\n",
       "        targ = e.target;\n",
       "    } else if (e.srcElement) {\n",
       "        targ = e.srcElement;\n",
       "    }\n",
       "    if (targ.nodeType === 3) {\n",
       "        // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "    }\n",
       "\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    var boundingRect = targ.getBoundingClientRect();\n",
       "    var x = e.pageX - (boundingRect.left + document.body.scrollLeft);\n",
       "    var y = e.pageY - (boundingRect.top + document.body.scrollTop);\n",
       "\n",
       "    return { x: x, y: y };\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys(original) {\n",
       "    return Object.keys(original).reduce(function (obj, key) {\n",
       "        if (typeof original[key] !== 'object') {\n",
       "            obj[key] = original[key];\n",
       "        }\n",
       "        return obj;\n",
       "    }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function (event, name) {\n",
       "    var canvas_pos = mpl.findpos(event);\n",
       "\n",
       "    if (name === 'button_press') {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * this.ratio;\n",
       "    var y = canvas_pos.y * this.ratio;\n",
       "\n",
       "    this.send_message(name, {\n",
       "        x: x,\n",
       "        y: y,\n",
       "        button: event.button,\n",
       "        step: event.step,\n",
       "        guiEvent: simpleKeys(event),\n",
       "    });\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (_event, _name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.key_event = function (event, name) {\n",
       "    // Prevent repeat events\n",
       "    if (name === 'key_press') {\n",
       "        if (event.which === this._key) {\n",
       "            return;\n",
       "        } else {\n",
       "            this._key = event.which;\n",
       "        }\n",
       "    }\n",
       "    if (name === 'key_release') {\n",
       "        this._key = null;\n",
       "    }\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which !== 17) {\n",
       "        value += 'ctrl+';\n",
       "    }\n",
       "    if (event.altKey && event.which !== 18) {\n",
       "        value += 'alt+';\n",
       "    }\n",
       "    if (event.shiftKey && event.which !== 16) {\n",
       "        value += 'shift+';\n",
       "    }\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, { key: value, guiEvent: simpleKeys(event) });\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function (name) {\n",
       "    if (name === 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message('toolbar_button', { name: name });\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function (tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "\n",
       "///////////////// REMAINING CONTENT GENERATED BY embed_js.py /////////////////\n",
       "// prettier-ignore\n",
       "var _JSXTOOLS_RESIZE_OBSERVER=function(A){var t,i=new WeakMap,n=new WeakMap,a=new WeakMap,r=new WeakMap,o=new Set;function s(e){if(!(this instanceof s))throw new TypeError(\"Constructor requires 'new' operator\");i.set(this,e)}function h(){throw new TypeError(\"Function is not a constructor\")}function c(e,t,i,n){e=0 in arguments?Number(arguments[0]):0,t=1 in arguments?Number(arguments[1]):0,i=2 in arguments?Number(arguments[2]):0,n=3 in arguments?Number(arguments[3]):0,this.right=(this.x=this.left=e)+(this.width=i),this.bottom=(this.y=this.top=t)+(this.height=n),Object.freeze(this)}function d(){t=requestAnimationFrame(d);var s=new WeakMap,p=new Set;o.forEach((function(t){r.get(t).forEach((function(i){var r=t instanceof window.SVGElement,o=a.get(t),d=r?0:parseFloat(o.paddingTop),f=r?0:parseFloat(o.paddingRight),l=r?0:parseFloat(o.paddingBottom),u=r?0:parseFloat(o.paddingLeft),g=r?0:parseFloat(o.borderTopWidth),m=r?0:parseFloat(o.borderRightWidth),w=r?0:parseFloat(o.borderBottomWidth),b=u+f,F=d+l,v=(r?0:parseFloat(o.borderLeftWidth))+m,W=g+w,y=r?0:t.offsetHeight-W-t.clientHeight,E=r?0:t.offsetWidth-v-t.clientWidth,R=b+v,z=F+W,M=r?t.width:parseFloat(o.width)-R-E,O=r?t.height:parseFloat(o.height)-z-y;if(n.has(t)){var k=n.get(t);if(k[0]===M&&k[1]===O)return}n.set(t,[M,O]);var S=Object.create(h.prototype);S.target=t,S.contentRect=new c(u,d,M,O),s.has(i)||(s.set(i,[]),p.add(i)),s.get(i).push(S)}))})),p.forEach((function(e){i.get(e).call(e,s.get(e),e)}))}return s.prototype.observe=function(i){if(i instanceof window.Element){r.has(i)||(r.set(i,new Set),o.add(i),a.set(i,window.getComputedStyle(i)));var n=r.get(i);n.has(this)||n.add(this),cancelAnimationFrame(t),t=requestAnimationFrame(d)}},s.prototype.unobserve=function(i){if(i instanceof window.Element&&r.has(i)){var n=r.get(i);n.has(this)&&(n.delete(this),n.size||(r.delete(i),o.delete(i))),n.size||r.delete(i),o.size||cancelAnimationFrame(t)}},A.DOMRectReadOnly=c,A.ResizeObserver=s,A.ResizeObserverEntry=h,A}; // eslint-disable-line\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Left button pans, Right button zooms\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";/* global mpl */\n",
       "\n",
       "var comm_websocket_adapter = function (comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function () {\n",
       "        comm.close();\n",
       "    };\n",
       "    ws.send = function (m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function (msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data']);\n",
       "    });\n",
       "    return ws;\n",
       "};\n",
       "\n",
       "mpl.mpl_figure_comm = function (comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = document.getElementById(id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm);\n",
       "\n",
       "    function ondownload(figure, _format) {\n",
       "        window.open(figure.canvas.toDataURL());\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy, ondownload, element);\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element;\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error('Failed to find cell for figure', id, fig);\n",
       "        return;\n",
       "    }\n",
       "    fig.cell_info[0].output_area.element.on(\n",
       "        'cleared',\n",
       "        { fig: fig },\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function (fig, msg) {\n",
       "    var width = fig.canvas.width / fig.ratio;\n",
       "    fig.cell_info[0].output_area.element.off(\n",
       "        'cleared',\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "    fig.resizeObserverInstance.unobserve(fig.canvas_div);\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable();\n",
       "    fig.parent_element.innerHTML =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "    fig.close_ws(fig, msg);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.close_ws = function (fig, msg) {\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function (_remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width / this.ratio;\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message('ack', {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () {\n",
       "        fig.push_to_output();\n",
       "    }, 1000);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'btn-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'btn-group';\n",
       "    var button;\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'btn-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        button = fig.buttons[name] = document.createElement('button');\n",
       "        button.classList = 'btn btn-default';\n",
       "        button.href = '#';\n",
       "        button.title = name;\n",
       "        button.innerHTML = '<i class=\"fa ' + image + ' fa-lg\"></i>';\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message pull-right';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = document.createElement('div');\n",
       "    buttongrp.classList = 'btn-group inline pull-right';\n",
       "    button = document.createElement('button');\n",
       "    button.classList = 'btn btn-mini btn-primary';\n",
       "    button.href = '#';\n",
       "    button.title = 'Stop Interaction';\n",
       "    button.innerHTML = '<i class=\"fa fa-power-off icon-remove icon-large\"></i>';\n",
       "    button.addEventListener('click', function (_evt) {\n",
       "        fig.handle_close(fig, {});\n",
       "    });\n",
       "    button.addEventListener(\n",
       "        'mouseover',\n",
       "        on_mouseover_closure('Stop Interaction')\n",
       "    );\n",
       "    buttongrp.appendChild(button);\n",
       "    var titlebar = this.root.querySelector('.ui-dialog-titlebar');\n",
       "    titlebar.insertBefore(buttongrp, titlebar.firstChild);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._remove_fig_handler = function (event) {\n",
       "    var fig = event.data.fig;\n",
       "    if (event.target !== this) {\n",
       "        // Ignore bubbled events from children.\n",
       "        return;\n",
       "    }\n",
       "    fig.close_ws(fig, {});\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (el) {\n",
       "    el.style.boxSizing = 'content-box'; // override notebook setting of border-box.\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (el) {\n",
       "    // this is important to make the div 'focusable\n",
       "    el.setAttribute('tabindex', 0);\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    } else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (event, _name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager) {\n",
       "        manager = IPython.keyboard_manager;\n",
       "    }\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which === 13) {\n",
       "        this.canvas_div.blur();\n",
       "        // select the cell after this one\n",
       "        var index = IPython.notebook.find_cell_index(this.cell_info[0]);\n",
       "        IPython.notebook.select(index + 1);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "};\n",
       "\n",
       "mpl.find_output_cell = function (html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i = 0; i < ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code') {\n",
       "            for (var j = 0; j < cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] === html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "};\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel !== null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target(\n",
       "        'matplotlib',\n",
       "        mpl.mpl_figure_comm\n",
       "    );\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"640\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x180ba409910>]"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%matplotlib notebook\n",
    "plt.plot(triple_barrier_events.index, triple_barrier_events['trgt'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "id": "corresponding-sport",
   "metadata": {},
   "outputs": [],
   "source": [
    "## https://ai.plainenglish.io/start-using-better-labels-for-financial-machine-learning-6eeac691e660"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "electrical-lover",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "id": "cubic-monitoring",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2021-12-23 15:11:45.872442 100.0% num_concurrent_events done after 0.11 minutes. Remaining 0.0 minutes..\n",
      "2021-12-23 15:11:52.012483 100.0% _get_average_uniqueness done after 0.1 minutes. Remaining 0.0 minutes.s.\n"
     ]
    }
   ],
   "source": [
    "from mlfinlab.sampling.concurrent import get_av_uniqueness_from_triple_barrier\n",
    "\n",
    "#barrier_events = pd.read_csv('NVAX_test.csv', index_col=0, parse_dates=[0,2])\n",
    "\n",
    "#close_prices = pd.read_csv('NVAX_test.csv', index_col=0, parse_dates=[0,2])\n",
    "\n",
    "av_unique = get_av_uniqueness_from_triple_barrier(triple_barrier_events, df['Close'],\n",
    "num_threads=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "id": "lesser-ivory",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>tW</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2020-02-03 17:31:00</th>\n",
       "      <td>0.592593</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-02-03 17:33:00</th>\n",
       "      <td>0.476190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-02-03 17:39:00</th>\n",
       "      <td>0.708333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-02-03 17:42:00</th>\n",
       "      <td>0.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-02-03 17:45:00</th>\n",
       "      <td>0.666667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                           tW\n",
       "2020-02-03 17:31:00  0.592593\n",
       "2020-02-03 17:33:00  0.476190\n",
       "2020-02-03 17:39:00  0.708333\n",
       "2020-02-03 17:42:00  0.750000\n",
       "2020-02-03 17:45:00  0.666667"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "av_unique.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "marked-circulation",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "id": "outdoor-liabilities",
   "metadata": {},
   "outputs": [],
   "source": [
    "stock = 'NVAX'\n",
    "df = pd.read_csv(r'C:\\Users\\Michael\\Desktop\\Python\\Stonks\\Data_Scrape\\file_storage_new\\NVAX.csv'\n",
    "                 , parse_dates = [['<DATE>', '<TIME>']])\n",
    "\n",
    "df.drop(['Unnamed: 0'], axis=1, inplace = True)\n",
    "\n",
    "\n",
    "df.columns = ['datetime','Open','High','Low','Close','Volume']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "nominated-friendly",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>datetime</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2020-01-31 17:30:00</td>\n",
       "      <td>7.2100</td>\n",
       "      <td>7.290</td>\n",
       "      <td>7.1500</td>\n",
       "      <td>7.1896</td>\n",
       "      <td>145419</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2020-01-31 17:31:00</td>\n",
       "      <td>7.2050</td>\n",
       "      <td>7.205</td>\n",
       "      <td>7.0000</td>\n",
       "      <td>7.1099</td>\n",
       "      <td>57453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2020-01-31 17:32:00</td>\n",
       "      <td>7.0300</td>\n",
       "      <td>7.110</td>\n",
       "      <td>7.0300</td>\n",
       "      <td>7.0300</td>\n",
       "      <td>22454</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2020-01-31 17:33:00</td>\n",
       "      <td>7.0300</td>\n",
       "      <td>7.060</td>\n",
       "      <td>6.9300</td>\n",
       "      <td>6.9400</td>\n",
       "      <td>60747</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2020-01-31 17:34:00</td>\n",
       "      <td>6.9301</td>\n",
       "      <td>7.050</td>\n",
       "      <td>6.9301</td>\n",
       "      <td>7.0400</td>\n",
       "      <td>25353</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             datetime    Open   High     Low   Close  Volume\n",
       "0 2020-01-31 17:30:00  7.2100  7.290  7.1500  7.1896  145419\n",
       "1 2020-01-31 17:31:00  7.2050  7.205  7.0000  7.1099   57453\n",
       "2 2020-01-31 17:32:00  7.0300  7.110  7.0300  7.0300   22454\n",
       "3 2020-01-31 17:33:00  7.0300  7.060  6.9300  6.9400   60747\n",
       "4 2020-01-31 17:34:00  6.9301  7.050  6.9301  7.0400   25353"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "expired-courtesy",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(175562, 6)"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "automated-legislature",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "df_ndaq = pd.read_csv(r'C:\\Users\\Michael\\Desktop\\Python\\Stonks\\Data_Scrape\\file_storage_new\\NASDAQ100_200131_211126.txt'\n",
    "                 , parse_dates = [['<DATE>', '<TIME>']])\n",
    "\n",
    "#df.drop(['Unnamed: 0'], axis=1, inplace = True)\n",
    "\n",
    "\n",
    "df_ndaq.columns = ['datetime','Open_ndaq','High_ndaq','Low_ndaq','Close_ndaq','Volume_ndaq']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "id": "final-friday",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>datetime</th>\n",
       "      <th>Open_ndaq</th>\n",
       "      <th>High_ndaq</th>\n",
       "      <th>Low_ndaq</th>\n",
       "      <th>Close_ndaq</th>\n",
       "      <th>Volume_ndaq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2020-01-31 00:01:00</td>\n",
       "      <td>9136.087</td>\n",
       "      <td>9136.087</td>\n",
       "      <td>9136.087</td>\n",
       "      <td>9136.087</td>\n",
       "      <td>2146163</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2020-01-31 17:32:00</td>\n",
       "      <td>9167.006</td>\n",
       "      <td>9167.006</td>\n",
       "      <td>9167.006</td>\n",
       "      <td>9167.006</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2020-01-31 17:33:00</td>\n",
       "      <td>9159.026</td>\n",
       "      <td>9159.026</td>\n",
       "      <td>9159.026</td>\n",
       "      <td>9159.026</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2020-01-31 17:34:00</td>\n",
       "      <td>9139.133</td>\n",
       "      <td>9139.133</td>\n",
       "      <td>9139.133</td>\n",
       "      <td>9139.133</td>\n",
       "      <td>982949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2020-01-31 17:35:00</td>\n",
       "      <td>9139.322</td>\n",
       "      <td>9139.322</td>\n",
       "      <td>9139.322</td>\n",
       "      <td>9139.322</td>\n",
       "      <td>989636</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             datetime  Open_ndaq  High_ndaq  Low_ndaq  Close_ndaq  Volume_ndaq\n",
       "0 2020-01-31 00:01:00   9136.087   9136.087  9136.087    9136.087      2146163\n",
       "1 2020-01-31 17:32:00   9167.006   9167.006  9167.006    9167.006            0\n",
       "2 2020-01-31 17:33:00   9159.026   9159.026  9159.026    9159.026            0\n",
       "3 2020-01-31 17:34:00   9139.133   9139.133  9139.133    9139.133       982949\n",
       "4 2020-01-31 17:35:00   9139.322   9139.322  9139.322    9139.322       989636"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_ndaq.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "id": "consistent-malawi",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(178256, 6)"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_ndaq.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "aboriginal-fraud",
   "metadata": {},
   "outputs": [],
   "source": [
    "combined_df = pd.merge(df, df_ndaq, on='datetime')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "id": "bearing-fiction",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>datetime</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "      <th>Open_ndaq</th>\n",
       "      <th>High_ndaq</th>\n",
       "      <th>Low_ndaq</th>\n",
       "      <th>Close_ndaq</th>\n",
       "      <th>Volume_ndaq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2020-01-31 17:32:00</td>\n",
       "      <td>7.0300</td>\n",
       "      <td>7.11</td>\n",
       "      <td>7.0300</td>\n",
       "      <td>7.0300</td>\n",
       "      <td>22454</td>\n",
       "      <td>9167.006</td>\n",
       "      <td>9167.006</td>\n",
       "      <td>9167.006</td>\n",
       "      <td>9167.006</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2020-01-31 17:33:00</td>\n",
       "      <td>7.0300</td>\n",
       "      <td>7.06</td>\n",
       "      <td>6.9300</td>\n",
       "      <td>6.9400</td>\n",
       "      <td>60747</td>\n",
       "      <td>9159.026</td>\n",
       "      <td>9159.026</td>\n",
       "      <td>9159.026</td>\n",
       "      <td>9159.026</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2020-01-31 17:34:00</td>\n",
       "      <td>6.9301</td>\n",
       "      <td>7.05</td>\n",
       "      <td>6.9301</td>\n",
       "      <td>7.0400</td>\n",
       "      <td>25353</td>\n",
       "      <td>9139.133</td>\n",
       "      <td>9139.133</td>\n",
       "      <td>9139.133</td>\n",
       "      <td>9139.133</td>\n",
       "      <td>982949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2020-01-31 17:35:00</td>\n",
       "      <td>7.0400</td>\n",
       "      <td>7.08</td>\n",
       "      <td>7.0000</td>\n",
       "      <td>7.0000</td>\n",
       "      <td>23505</td>\n",
       "      <td>9139.322</td>\n",
       "      <td>9139.322</td>\n",
       "      <td>9139.322</td>\n",
       "      <td>9139.322</td>\n",
       "      <td>989636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2020-01-31 17:36:00</td>\n",
       "      <td>6.9700</td>\n",
       "      <td>7.02</td>\n",
       "      <td>6.9500</td>\n",
       "      <td>7.0173</td>\n",
       "      <td>12454</td>\n",
       "      <td>9141.547</td>\n",
       "      <td>9141.547</td>\n",
       "      <td>9141.547</td>\n",
       "      <td>9141.547</td>\n",
       "      <td>751910</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             datetime    Open  High     Low   Close  Volume  Open_ndaq  \\\n",
       "0 2020-01-31 17:32:00  7.0300  7.11  7.0300  7.0300   22454   9167.006   \n",
       "1 2020-01-31 17:33:00  7.0300  7.06  6.9300  6.9400   60747   9159.026   \n",
       "2 2020-01-31 17:34:00  6.9301  7.05  6.9301  7.0400   25353   9139.133   \n",
       "3 2020-01-31 17:35:00  7.0400  7.08  7.0000  7.0000   23505   9139.322   \n",
       "4 2020-01-31 17:36:00  6.9700  7.02  6.9500  7.0173   12454   9141.547   \n",
       "\n",
       "   High_ndaq  Low_ndaq  Close_ndaq  Volume_ndaq  \n",
       "0   9167.006  9167.006    9167.006            0  \n",
       "1   9159.026  9159.026    9159.026            0  \n",
       "2   9139.133  9139.133    9139.133       982949  \n",
       "3   9139.322  9139.322    9139.322       989636  \n",
       "4   9141.547  9141.547    9141.547       751910  "
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "combined_df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "id": "coral-folks",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(173648, 11)"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "combined_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "id": "respective-pointer",
   "metadata": {},
   "outputs": [],
   "source": [
    "combined_df = combined_df.set_index(['datetime'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "id": "large-sister",
   "metadata": {},
   "outputs": [],
   "source": [
    "import scipy.stats as stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "id": "quick-sensitivity",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pearson r: 0.860994989891818 and p-value: 0.0\n"
     ]
    }
   ],
   "source": [
    "r, p = stats.pearsonr(combined_df['Close'], combined_df['Close_ndaq'])\n",
    "\n",
    "print(f\"Pearson r: {r} and p-value: {p}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "id": "fallen-cuisine",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                      Close  Close_ndaq\n",
      "datetime                               \n",
      "2020-01-31 17:32:00  7.0300    9167.006\n",
      "2020-01-31 17:33:00  6.9400    9159.026\n",
      "2020-01-31 17:34:00  7.0400    9139.133\n",
      "2020-01-31 17:35:00  7.0000    9139.322\n",
      "2020-01-31 17:36:00  7.0173    9141.547\n"
     ]
    }
   ],
   "source": [
    "new_df = combined_df[['Close', 'Close_ndaq']]\n",
    "\n",
    "print(new_df.head(5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "recent-qatar",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "id": "joined-pittsburgh",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "/* global mpl */\n",
       "window.mpl = {};\n",
       "\n",
       "mpl.get_websocket_type = function () {\n",
       "    if (typeof WebSocket !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof MozWebSocket !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert(\n",
       "            'Your browser does not have WebSocket support. ' +\n",
       "                'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "                'Firefox 4 and 5 are also supported but you ' +\n",
       "                'have to enable WebSockets in about:config.'\n",
       "        );\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure = function (figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = this.ws.binaryType !== undefined;\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById('mpl-warnings');\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent =\n",
       "                'This browser does not support binary websocket messages. ' +\n",
       "                'Performance may be slow.';\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = document.createElement('div');\n",
       "    this.root.setAttribute('style', 'display: inline-block');\n",
       "    this._root_extra_style(this.root);\n",
       "\n",
       "    parent_element.appendChild(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen = function () {\n",
       "        fig.send_message('supports_binary', { value: fig.supports_binary });\n",
       "        fig.send_message('send_image_mode', {});\n",
       "        if (fig.ratio !== 1) {\n",
       "            fig.send_message('set_dpi_ratio', { dpi_ratio: fig.ratio });\n",
       "        }\n",
       "        fig.send_message('refresh', {});\n",
       "    };\n",
       "\n",
       "    this.imageObj.onload = function () {\n",
       "        if (fig.image_mode === 'full') {\n",
       "            // Full images could contain transparency (where diff images\n",
       "            // almost always do), so we need to clear the canvas so that\n",
       "            // there is no ghosting.\n",
       "            fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "        }\n",
       "        fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "    };\n",
       "\n",
       "    this.imageObj.onunload = function () {\n",
       "        fig.ws.close();\n",
       "    };\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_header = function () {\n",
       "    var titlebar = document.createElement('div');\n",
       "    titlebar.classList =\n",
       "        'ui-dialog-titlebar ui-widget-header ui-corner-all ui-helper-clearfix';\n",
       "    var titletext = document.createElement('div');\n",
       "    titletext.classList = 'ui-dialog-title';\n",
       "    titletext.setAttribute(\n",
       "        'style',\n",
       "        'width: 100%; text-align: center; padding: 3px;'\n",
       "    );\n",
       "    titlebar.appendChild(titletext);\n",
       "    this.root.appendChild(titlebar);\n",
       "    this.header = titletext;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = (this.canvas_div = document.createElement('div'));\n",
       "    canvas_div.setAttribute(\n",
       "        'style',\n",
       "        'border: 1px solid #ddd;' +\n",
       "            'box-sizing: content-box;' +\n",
       "            'clear: both;' +\n",
       "            'min-height: 1px;' +\n",
       "            'min-width: 1px;' +\n",
       "            'outline: 0;' +\n",
       "            'overflow: hidden;' +\n",
       "            'position: relative;' +\n",
       "            'resize: both;'\n",
       "    );\n",
       "\n",
       "    function on_keyboard_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.key_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    canvas_div.addEventListener(\n",
       "        'keydown',\n",
       "        on_keyboard_event_closure('key_press')\n",
       "    );\n",
       "    canvas_div.addEventListener(\n",
       "        'keyup',\n",
       "        on_keyboard_event_closure('key_release')\n",
       "    );\n",
       "\n",
       "    this._canvas_extra_style(canvas_div);\n",
       "    this.root.appendChild(canvas_div);\n",
       "\n",
       "    var canvas = (this.canvas = document.createElement('canvas'));\n",
       "    canvas.classList.add('mpl-canvas');\n",
       "    canvas.setAttribute('style', 'box-sizing: content-box;');\n",
       "\n",
       "    this.context = canvas.getContext('2d');\n",
       "\n",
       "    var backingStore =\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        this.context.webkitBackingStorePixelRatio ||\n",
       "        this.context.mozBackingStorePixelRatio ||\n",
       "        this.context.msBackingStorePixelRatio ||\n",
       "        this.context.oBackingStorePixelRatio ||\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        1;\n",
       "\n",
       "    this.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband_canvas = (this.rubberband_canvas = document.createElement(\n",
       "        'canvas'\n",
       "    ));\n",
       "    rubberband_canvas.setAttribute(\n",
       "        'style',\n",
       "        'box-sizing: content-box; position: absolute; left: 0; top: 0; z-index: 1;'\n",
       "    );\n",
       "\n",
       "    // Apply a ponyfill if ResizeObserver is not implemented by browser.\n",
       "    if (this.ResizeObserver === undefined) {\n",
       "        if (window.ResizeObserver !== undefined) {\n",
       "            this.ResizeObserver = window.ResizeObserver;\n",
       "        } else {\n",
       "            var obs = _JSXTOOLS_RESIZE_OBSERVER({});\n",
       "            this.ResizeObserver = obs.ResizeObserver;\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.resizeObserverInstance = new this.ResizeObserver(function (entries) {\n",
       "        var nentries = entries.length;\n",
       "        for (var i = 0; i < nentries; i++) {\n",
       "            var entry = entries[i];\n",
       "            var width, height;\n",
       "            if (entry.contentBoxSize) {\n",
       "                if (entry.contentBoxSize instanceof Array) {\n",
       "                    // Chrome 84 implements new version of spec.\n",
       "                    width = entry.contentBoxSize[0].inlineSize;\n",
       "                    height = entry.contentBoxSize[0].blockSize;\n",
       "                } else {\n",
       "                    // Firefox implements old version of spec.\n",
       "                    width = entry.contentBoxSize.inlineSize;\n",
       "                    height = entry.contentBoxSize.blockSize;\n",
       "                }\n",
       "            } else {\n",
       "                // Chrome <84 implements even older version of spec.\n",
       "                width = entry.contentRect.width;\n",
       "                height = entry.contentRect.height;\n",
       "            }\n",
       "\n",
       "            // Keep the size of the canvas and rubber band canvas in sync with\n",
       "            // the canvas container.\n",
       "            if (entry.devicePixelContentBoxSize) {\n",
       "                // Chrome 84 implements new version of spec.\n",
       "                canvas.setAttribute(\n",
       "                    'width',\n",
       "                    entry.devicePixelContentBoxSize[0].inlineSize\n",
       "                );\n",
       "                canvas.setAttribute(\n",
       "                    'height',\n",
       "                    entry.devicePixelContentBoxSize[0].blockSize\n",
       "                );\n",
       "            } else {\n",
       "                canvas.setAttribute('width', width * fig.ratio);\n",
       "                canvas.setAttribute('height', height * fig.ratio);\n",
       "            }\n",
       "            canvas.setAttribute(\n",
       "                'style',\n",
       "                'width: ' + width + 'px; height: ' + height + 'px;'\n",
       "            );\n",
       "\n",
       "            rubberband_canvas.setAttribute('width', width);\n",
       "            rubberband_canvas.setAttribute('height', height);\n",
       "\n",
       "            // And update the size in Python. We ignore the initial 0/0 size\n",
       "            // that occurs as the element is placed into the DOM, which should\n",
       "            // otherwise not happen due to the minimum size styling.\n",
       "            if (fig.ws.readyState == 1 && width != 0 && height != 0) {\n",
       "                fig.request_resize(width, height);\n",
       "            }\n",
       "        }\n",
       "    });\n",
       "    this.resizeObserverInstance.observe(canvas_div);\n",
       "\n",
       "    function on_mouse_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.mouse_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousedown',\n",
       "        on_mouse_event_closure('button_press')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseup',\n",
       "        on_mouse_event_closure('button_release')\n",
       "    );\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousemove',\n",
       "        on_mouse_event_closure('motion_notify')\n",
       "    );\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseenter',\n",
       "        on_mouse_event_closure('figure_enter')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseleave',\n",
       "        on_mouse_event_closure('figure_leave')\n",
       "    );\n",
       "\n",
       "    canvas_div.addEventListener('wheel', function (event) {\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        on_mouse_event_closure('scroll')(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.appendChild(canvas);\n",
       "    canvas_div.appendChild(rubberband_canvas);\n",
       "\n",
       "    this.rubberband_context = rubberband_canvas.getContext('2d');\n",
       "    this.rubberband_context.strokeStyle = '#000000';\n",
       "\n",
       "    this._resize_canvas = function (width, height, forward) {\n",
       "        if (forward) {\n",
       "            canvas_div.style.width = width + 'px';\n",
       "            canvas_div.style.height = height + 'px';\n",
       "        }\n",
       "    };\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    this.rubberband_canvas.addEventListener('contextmenu', function (_e) {\n",
       "        event.preventDefault();\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus() {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'mpl-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'mpl-button-group';\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'mpl-button-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        var button = (fig.buttons[name] = document.createElement('button'));\n",
       "        button.classList = 'mpl-widget';\n",
       "        button.setAttribute('role', 'button');\n",
       "        button.setAttribute('aria-disabled', 'false');\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "\n",
       "        var icon_img = document.createElement('img');\n",
       "        icon_img.src = '_images/' + image + '.png';\n",
       "        icon_img.srcset = '_images/' + image + '_large.png 2x';\n",
       "        icon_img.alt = tooltip;\n",
       "        button.appendChild(icon_img);\n",
       "\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    var fmt_picker = document.createElement('select');\n",
       "    fmt_picker.classList = 'mpl-widget';\n",
       "    toolbar.appendChild(fmt_picker);\n",
       "    this.format_dropdown = fmt_picker;\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = document.createElement('option');\n",
       "        option.selected = fmt === mpl.default_extension;\n",
       "        option.innerHTML = fmt;\n",
       "        fmt_picker.appendChild(option);\n",
       "    }\n",
       "\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.request_resize = function (x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', { width: x_pixels, height: y_pixels });\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_message = function (type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function () {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({ type: 'draw', figure_id: this.id }));\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function (fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] !== fig.canvas.width || size[1] !== fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1], msg['forward']);\n",
       "        fig.send_message('refresh', {});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function (fig, msg) {\n",
       "    var x0 = msg['x0'] / fig.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / fig.ratio;\n",
       "    var x1 = msg['x1'] / fig.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / fig.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0,\n",
       "        0,\n",
       "        fig.canvas.width / fig.ratio,\n",
       "        fig.canvas.height / fig.ratio\n",
       "    );\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function (fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function (fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch (cursor) {\n",
       "        case 0:\n",
       "            cursor = 'pointer';\n",
       "            break;\n",
       "        case 1:\n",
       "            cursor = 'default';\n",
       "            break;\n",
       "        case 2:\n",
       "            cursor = 'crosshair';\n",
       "            break;\n",
       "        case 3:\n",
       "            cursor = 'move';\n",
       "            break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_message = function (fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function (fig, _msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function (fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_history_buttons = function (fig, msg) {\n",
       "    for (var key in msg) {\n",
       "        if (!(key in fig.buttons)) {\n",
       "            continue;\n",
       "        }\n",
       "        fig.buttons[key].disabled = !msg[key];\n",
       "        fig.buttons[key].setAttribute('aria-disabled', !msg[key]);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_navigate_mode = function (fig, msg) {\n",
       "    if (msg['mode'] === 'PAN') {\n",
       "        fig.buttons['Pan'].classList.add('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    } else if (msg['mode'] === 'ZOOM') {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.add('active');\n",
       "    } else {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message('ack', {});\n",
       "};\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function (fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = 'image/png';\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src\n",
       "                );\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data\n",
       "            );\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        } else if (\n",
       "            typeof evt.data === 'string' &&\n",
       "            evt.data.slice(0, 21) === 'data:image/png;base64'\n",
       "        ) {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig['handle_' + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\n",
       "                \"No handler for the '\" + msg_type + \"' message type: \",\n",
       "                msg\n",
       "            );\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\n",
       "                    \"Exception inside the 'handler_\" + msg_type + \"' callback:\",\n",
       "                    e,\n",
       "                    e.stack,\n",
       "                    msg\n",
       "                );\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "};\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function (e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e) {\n",
       "        e = window.event;\n",
       "    }\n",
       "    if (e.target) {\n",
       "        targ = e.target;\n",
       "    } else if (e.srcElement) {\n",
       "        targ = e.srcElement;\n",
       "    }\n",
       "    if (targ.nodeType === 3) {\n",
       "        // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "    }\n",
       "\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    var boundingRect = targ.getBoundingClientRect();\n",
       "    var x = e.pageX - (boundingRect.left + document.body.scrollLeft);\n",
       "    var y = e.pageY - (boundingRect.top + document.body.scrollTop);\n",
       "\n",
       "    return { x: x, y: y };\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys(original) {\n",
       "    return Object.keys(original).reduce(function (obj, key) {\n",
       "        if (typeof original[key] !== 'object') {\n",
       "            obj[key] = original[key];\n",
       "        }\n",
       "        return obj;\n",
       "    }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function (event, name) {\n",
       "    var canvas_pos = mpl.findpos(event);\n",
       "\n",
       "    if (name === 'button_press') {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * this.ratio;\n",
       "    var y = canvas_pos.y * this.ratio;\n",
       "\n",
       "    this.send_message(name, {\n",
       "        x: x,\n",
       "        y: y,\n",
       "        button: event.button,\n",
       "        step: event.step,\n",
       "        guiEvent: simpleKeys(event),\n",
       "    });\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (_event, _name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.key_event = function (event, name) {\n",
       "    // Prevent repeat events\n",
       "    if (name === 'key_press') {\n",
       "        if (event.which === this._key) {\n",
       "            return;\n",
       "        } else {\n",
       "            this._key = event.which;\n",
       "        }\n",
       "    }\n",
       "    if (name === 'key_release') {\n",
       "        this._key = null;\n",
       "    }\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which !== 17) {\n",
       "        value += 'ctrl+';\n",
       "    }\n",
       "    if (event.altKey && event.which !== 18) {\n",
       "        value += 'alt+';\n",
       "    }\n",
       "    if (event.shiftKey && event.which !== 16) {\n",
       "        value += 'shift+';\n",
       "    }\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, { key: value, guiEvent: simpleKeys(event) });\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function (name) {\n",
       "    if (name === 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message('toolbar_button', { name: name });\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function (tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "\n",
       "///////////////// REMAINING CONTENT GENERATED BY embed_js.py /////////////////\n",
       "// prettier-ignore\n",
       "var _JSXTOOLS_RESIZE_OBSERVER=function(A){var t,i=new WeakMap,n=new WeakMap,a=new WeakMap,r=new WeakMap,o=new Set;function s(e){if(!(this instanceof s))throw new TypeError(\"Constructor requires 'new' operator\");i.set(this,e)}function h(){throw new TypeError(\"Function is not a constructor\")}function c(e,t,i,n){e=0 in arguments?Number(arguments[0]):0,t=1 in arguments?Number(arguments[1]):0,i=2 in arguments?Number(arguments[2]):0,n=3 in arguments?Number(arguments[3]):0,this.right=(this.x=this.left=e)+(this.width=i),this.bottom=(this.y=this.top=t)+(this.height=n),Object.freeze(this)}function d(){t=requestAnimationFrame(d);var s=new WeakMap,p=new Set;o.forEach((function(t){r.get(t).forEach((function(i){var r=t instanceof window.SVGElement,o=a.get(t),d=r?0:parseFloat(o.paddingTop),f=r?0:parseFloat(o.paddingRight),l=r?0:parseFloat(o.paddingBottom),u=r?0:parseFloat(o.paddingLeft),g=r?0:parseFloat(o.borderTopWidth),m=r?0:parseFloat(o.borderRightWidth),w=r?0:parseFloat(o.borderBottomWidth),b=u+f,F=d+l,v=(r?0:parseFloat(o.borderLeftWidth))+m,W=g+w,y=r?0:t.offsetHeight-W-t.clientHeight,E=r?0:t.offsetWidth-v-t.clientWidth,R=b+v,z=F+W,M=r?t.width:parseFloat(o.width)-R-E,O=r?t.height:parseFloat(o.height)-z-y;if(n.has(t)){var k=n.get(t);if(k[0]===M&&k[1]===O)return}n.set(t,[M,O]);var S=Object.create(h.prototype);S.target=t,S.contentRect=new c(u,d,M,O),s.has(i)||(s.set(i,[]),p.add(i)),s.get(i).push(S)}))})),p.forEach((function(e){i.get(e).call(e,s.get(e),e)}))}return s.prototype.observe=function(i){if(i instanceof window.Element){r.has(i)||(r.set(i,new Set),o.add(i),a.set(i,window.getComputedStyle(i)));var n=r.get(i);n.has(this)||n.add(this),cancelAnimationFrame(t),t=requestAnimationFrame(d)}},s.prototype.unobserve=function(i){if(i instanceof window.Element&&r.has(i)){var n=r.get(i);n.has(this)&&(n.delete(this),n.size||(r.delete(i),o.delete(i))),n.size||r.delete(i),o.size||cancelAnimationFrame(t)}},A.DOMRectReadOnly=c,A.ResizeObserver=s,A.ResizeObserverEntry=h,A}; // eslint-disable-line\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Left button pans, Right button zooms\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";/* global mpl */\n",
       "\n",
       "var comm_websocket_adapter = function (comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function () {\n",
       "        comm.close();\n",
       "    };\n",
       "    ws.send = function (m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function (msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data']);\n",
       "    });\n",
       "    return ws;\n",
       "};\n",
       "\n",
       "mpl.mpl_figure_comm = function (comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = document.getElementById(id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm);\n",
       "\n",
       "    function ondownload(figure, _format) {\n",
       "        window.open(figure.canvas.toDataURL());\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy, ondownload, element);\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element;\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error('Failed to find cell for figure', id, fig);\n",
       "        return;\n",
       "    }\n",
       "    fig.cell_info[0].output_area.element.on(\n",
       "        'cleared',\n",
       "        { fig: fig },\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function (fig, msg) {\n",
       "    var width = fig.canvas.width / fig.ratio;\n",
       "    fig.cell_info[0].output_area.element.off(\n",
       "        'cleared',\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "    fig.resizeObserverInstance.unobserve(fig.canvas_div);\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable();\n",
       "    fig.parent_element.innerHTML =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "    fig.close_ws(fig, msg);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.close_ws = function (fig, msg) {\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function (_remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width / this.ratio;\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message('ack', {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () {\n",
       "        fig.push_to_output();\n",
       "    }, 1000);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'btn-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'btn-group';\n",
       "    var button;\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'btn-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        button = fig.buttons[name] = document.createElement('button');\n",
       "        button.classList = 'btn btn-default';\n",
       "        button.href = '#';\n",
       "        button.title = name;\n",
       "        button.innerHTML = '<i class=\"fa ' + image + ' fa-lg\"></i>';\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message pull-right';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = document.createElement('div');\n",
       "    buttongrp.classList = 'btn-group inline pull-right';\n",
       "    button = document.createElement('button');\n",
       "    button.classList = 'btn btn-mini btn-primary';\n",
       "    button.href = '#';\n",
       "    button.title = 'Stop Interaction';\n",
       "    button.innerHTML = '<i class=\"fa fa-power-off icon-remove icon-large\"></i>';\n",
       "    button.addEventListener('click', function (_evt) {\n",
       "        fig.handle_close(fig, {});\n",
       "    });\n",
       "    button.addEventListener(\n",
       "        'mouseover',\n",
       "        on_mouseover_closure('Stop Interaction')\n",
       "    );\n",
       "    buttongrp.appendChild(button);\n",
       "    var titlebar = this.root.querySelector('.ui-dialog-titlebar');\n",
       "    titlebar.insertBefore(buttongrp, titlebar.firstChild);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._remove_fig_handler = function (event) {\n",
       "    var fig = event.data.fig;\n",
       "    if (event.target !== this) {\n",
       "        // Ignore bubbled events from children.\n",
       "        return;\n",
       "    }\n",
       "    fig.close_ws(fig, {});\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (el) {\n",
       "    el.style.boxSizing = 'content-box'; // override notebook setting of border-box.\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (el) {\n",
       "    // this is important to make the div 'focusable\n",
       "    el.setAttribute('tabindex', 0);\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    } else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (event, _name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager) {\n",
       "        manager = IPython.keyboard_manager;\n",
       "    }\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which === 13) {\n",
       "        this.canvas_div.blur();\n",
       "        // select the cell after this one\n",
       "        var index = IPython.notebook.find_cell_index(this.cell_info[0]);\n",
       "        IPython.notebook.select(index + 1);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "};\n",
       "\n",
       "mpl.find_output_cell = function (html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i = 0; i < ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code') {\n",
       "            for (var j = 0; j < cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] === html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "};\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel !== null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target(\n",
       "        'matplotlib',\n",
       "        mpl.mpl_figure_comm\n",
       "    );\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAArwAAAEsCAYAAAAhNGCdAAAgAElEQVR4nOzdeXxMV+MG8BtLQqSJKCUNgtqKtnRRvLroolpedP/RKlq6v1WtFrXEVlvt+05Rse8iyCIkIajIHtkJWRBJRPZknt8f1yw3M3eyzUwyk+f7+ZyPzN3m3OkpT86ce44AIiIiIiILJlR3BYiIiIiIjImBl4iIiIgsGgMvEREREVk0Bl4iIiIismgMvERERERk0Rh4iYiIiMiiMfASERERkUVj4CUiIiIii8bAS0REREQWjYGXiIiIiCwaAy8RERERWTQGXiIiIiKyaAy8RERERGTRGHiJyKL5+PhAEAT4+Pioto0cORIuLi7VViciIjItBl4i0issLAyfffYZnnzySVhbW8PJyQnDhw9HWFhYdVetXKoSeF977TUIgqAqjo6OePHFF7F582aUlJQYr9JUpjVr1uCjjz5Cq1atIAgCRo4cWaHzS0pKsGDBArRp0wY2NjZ45plnsGvXLuNUloiqHQMvEck6cOAArK2t0aJFC0yZMgWbNm3C1KlT4eTkBGtraxw8eLC6q1imqgbeli1bYseOHdixYweWLFmC7t27QxAETJw40XiVpjK5uLigSZMmGDBgAOrVq1fhwDtp0iQIgoCxY8diw4YNGDhwIARBgJubm3EqTETVioGXiHSKjY2Fra0tOnfujDt37kj23b17F507d0ajRo0QFxdn0no9fPiwQsdXNfB27dpVsi0nJwctW7ZEo0aNUFhYWKG6VFZJSQny8vJM8l5VUdH/NlWRmJgIhUIBAGjUqFGFAu+tW7dQv359/PDDD6ptCoUCr7zyClq2bIni4mJDV5eIqhkDLxHp9M0330AQBJw7d07nfl9fXwiCgG+++QYAsG/fPgiCgLNnz2odu27dOgiCgNDQUNW2yMhIfPjhh3B0dISNjQ1eeOEFHDlyRHLe1q1bVdf87rvv0KxZMzRu3BiAGHi+++47dOzYEQ0aNECTJk3w0UcfISEhQXINQwdeAPjoo48gCAJu374NAMjIyMC4cePQsmVLWFtb46mnnsL8+fO1hj389ddf6N27N5o0aYIGDRrg+eefx759+7SuLwgCfvjhB+zcuRNdunRBvXr1cOjQIQCAm5sbnn/+edjZ2eGxxx5Dt27dsGzZMsn5cXFx+Oijj+Do6IiGDRvi5ZdfxvHjx3V+Lnv27MGcOXPg7OwMGxsbvPHGG4iJiSnzs3F1dYUgCAgPD8ewYcPQuHFjdO/evczzjKGigXf16tWqumvatWsXBEHA+fPnDVxDIqpuDLxEpNOTTz6JNm3a6D2mTZs2aNmyJQAgNzcXdnZ2+P7777WO69evnyQ4hoWFwcHBAV26dMGCBQuwatUqvPrqq7CyspIMk1AG3i5duuC1117DypUrMX/+fABiwH7uuecwffp0bNiwAX/88QccHR3h4uKCnJwc1TWMEXiff/551K1bFzk5OcjJycGzzz6Lxx9/HH/88QfWrVuHL774AlZWVhg3bpzkvJYtW+L777/HqlWrsGTJEvTs2ROCIGiFUUEQ8PTTT6NZs2aYOXMmVq9ejaCgIJw+fRqCIODNN9/E6tWrsXr1avz444/4+OOPVeempqaiefPmeOyxxzBlyhQsWbIEzz33HOrUqSP5bJWfS48ePfDCCy9g6dKlmDFjBmxtbdGzZ88yPxtl4O3SpQuGDBmCNWvWYPXq1bLHl5SU4O7du+UqFe05r2jgHTNmDBo1aqTqIVaKjY2FIAhYsWJFhd6fiGo+Bl4i0pKZmQlBEDBkyBC9xw0ePBiCIODBgwcAgGHDhuGJJ56QfCWckpKCOnXqYNasWaptb775Jp555hnk5+ertikUCvTp0wcdOnRQbVMG3r59+2p9zZybm6tVnwsXLkAQBGzfvl21raqBt3PnzqogFhkZiZ9++gmCIOC///0vAGD27Nlo1KgRoqOjJedOmjQJdevWxc2bN2XrXFhYiG7duuGNN96QbBcEAXXq1NHqgRw3bhzs7e31fuX+888/a/VSZmdno23btmjTpo2q11n5uTz99NMoKChQHbt8+XKt3nhdlIF32LBheo9TSkhIkDwAqK9o/rcqj4oG3oEDB6Jdu3Za23NyciAIAiZNmlSh9yeimo+Bl4i0JCUlQRAEfP7553qP++yzzyAIAm7dugUAOHz4MARBgKenp+qYlStXQhAEXL9+HQCQnp4OKysrzJ49W6tnb+bMmZLrKQPv33//rbcehYWFuHfvHu7evYvGjRvj559/Vu0z5CwNgiDAysoKAwcOxN27dwEAzz77LAYMGKB1L56enhAEATt37tR57fv37+Pu3bv47rvvVMM0lARBQL9+/bTOcXV1Rd26dXHy5EnZOnfs2FFnD+28efMkQVb5uSxcuFBy3NWrVyEIgtbwEl11EQQBvr6+eo9TysvLw5kzZ8pV7t+/X65rKlU08L7xxht4+umntbaXlJRAEAStnnkiMn8MvESkpaI9vFlZWQCA/Px8ODg4YOzYsapj+vbtKxnbGRgYWGYP39WrVwGoA6+uccS5ubmYNm0aWrZsCSsrK8n5o0ePVh1X1cDbpk0bnDlzBp6envDz80NaWprkmIYNG+q9lyVLlqiOPXbsGF5++WXY2NhohWhNgiDgyy+/1KpPWloann76aQiCAGdnZ4wePVor/NrY2GDEiBFa5yp/GVEOn1B+Lrt375Ycp+yJ3bZtm97PRhl4NXuwqwt7eImoLAy8RKSTk5MT2rZtq/eYNm3awNnZWbJt5MiRaNq0KYqKinDr1i1YWVlh3rx5qv3KYQcTJkyQ7eFTDpFQBt7Lly9rvfdXX32FOnXq4JdffsG+fftw+vRpnDlzBo8//rgk/BhjDK8mGxsbvP3227L3cuPGDQDAuXPnYGVlhddeew2bN2+Gu7s7zpw5g+HDh0MQpH8VKx9a06WgoABHjx7Fd999hzZt2kAQBHzxxReS+lQk8JZ+aE4ZeLdu3ar3vpWBV9nTXZbi4mKkpKSUq2gOsSiPyozhtbW15RheolqEgZeIdBo7dqzeJ9bPnTsnmaVByd3dHYIgwMPDA0uXLoUgCIiPj1ftT0tLgyAImDx5cpl10Bd4HRwcJD25gPi1ed26dU0aeLt06YLevXuXea1x48ahYcOGknHLACoceDWVlJSoZtNQzqwgN6Rh/vz5Ooc0mCrw1qQxvKtWrdI5S8M///yjd2YSIjJfDLxEpFN0dDQaNmyILl264N69e5J96enp6NKlC2xtbREbGyvZV1hYiCZNmmD06NHo1auXzvD1+uuvo0mTJkhOTtbapznnr77A26RJE4waNUqybeHChRAEwaSBd8aMGaqAX1pGRgaKiooAAL/88gtsbW0lM0gkJCTA1ta23IG39H8HQD3FlnLlO+VDawEBAapjHj58iHbt2ul8aM1Ugbe6xvBmZmYiMjISmZmZqm1JSUmy8/A6OztzHl4iC8TAS0Sy9u7di/r168PJyQlTp07F5s2bMW3aNNUywwcOHNB53pgxY2BnZwcrKyssXrxYa394eDgcHR3x+OOPY9KkSdiwYQNmz56N9957D88++6zqOH2B94svvkDdunUxbtw4rF+/HqNGjULLli1NPqQhJycHzz//POrVq4cxY8Zg7dq1WLRoEUaOHIlGjRqpAqGXlxcEQcArr7yCtWvXYubMmXjiiSfw7LPPljvwDh06FK+++ipmzJiBTZs2Ydq0aar5b5VBVjktmYODA6ZNm4alS5eie/fuWlO+mTrwGtrRo0cxe/ZszJ49G9bW1ujRo4fqdXBwsOo4ZRsqfT+//fYbBEHA119/jY0bN6pWWvvnn39MfCdEZAoMvESkV0hICIYNGwYnJyfUr18fLVq0wLBhw/ROW3XmzBnVw1hJSUk6j4mLi8MXX3yBFi1aoH79+nB2dsagQYOwf/9+1TH6Am9GRgZGjx6Npk2bws7ODu+88w6ioqLg4uJi0sALiNN+TZ48Ge3bt4e1tTWaNm2KPn36YNGiRZI5ZTdv3owOHTrAxsYGnTt3xtatW1XBUZNc4N2/fz/69++PJ554AtbW1mjdujW++eYbpKSkSI5TLjzRuHFjNGjQAD179pRdeMJcA+/IkSNlh0Ro1l0u8JaUlGDu3LlwcXGBtbU1unbtKjujBhGZPwZeIiIiIrJoDLxEREREZNEYeImIiIjIojHwEhEREZFFY+AlIiIiIovGwEtEREREFo2Bl4iIiIgsGgOvEZWUlCApKQmZmZnIyspiYWFhYWFhYTFayczMRFJSkmohGlJj4DWipKSkcq8dz8LCwsLCwsJiiCK34E9txsBrRJmZmaqGV92/9bGwsLCwsLBYdlF2tGVmZlZ3BKpxGHiNKCsrC4IgICsrq7qrQkRERBaOuUMeA68RseERERGRqTB3yGPgNSI2PCIiIjIV5g55DLxGxIZHREREpsLcIY+B14jK0/CKi4uRl5fHUgsLp40hIiJDYuCVx8BrRPoankKhQHJyMiIiIlhqaYmKikJBQUE1tEwiIrJEDLzyGHiNSF/DU4bde/fuITc3t9p7G1lMW3JychATE4PExEQoFIpqaJ1ERGR00aeBvaOA3PsmeTsGXnkMvEYk1/CKi4tVYZdqr8zMTERERKCwsLC6q0JERIaUfQdICQVc7cVy/FeTvC0DrzwGXiOSa3h5eXmIiIhAbm5uNdWMaoLc3FxEREQgLy+vuqtCRESGpAy6yrLieZO8LQOvPAZeIyor8DLo1G5sB0REZkShAKLcgQcp+o87/L124HW1N0kVGXjlMfAaEQMv6cN2QERkRrYNUofXuS3FP5MuS4/JSZeG3Pku6p+DdgF5xl3yl4FXHgOvEdXmwCsIAg4dOlTd1ajRakM7ICKyCEUFunttNXtui4u09wX9I339pzNwP9Fo1WTglcfAa0SWHHhTUlLw448/om3btrC2tkbLli0xaNAgeHp6AmDgLQ9LaAdERLVCepxG724r+fCrWTKTxHNLb7/uYbRqMvDKY+A1IksNvAkJCXjyySfRpUsX7N+/H9evX0dYWBgWL16MTp06AWDgLQ9zbwdERLXGga/FsDqvlfg6+rT+sKup9L4TE4xWTQZeeQy8RmSpgffdd9+Fs7MzHj58qLUvIyMDgHbgDQkJQb9+/dCgQQM0adIEY8eORXZ2tmq/j48PXnrpJdja2sLBwQF9+vRBYqL6a5/Dhw+jR48esLGxQdu2bTFjxgwUFRUZ8S6Nz9zbARFRrVBSog6rfw9Wby8dZLcOFP+M9ZaeH+MpPe7I/4xWVQZeeQy8RlSRwKtQKJBTUFQtpSILH6Snp8PKygpz587Ve5xm4H348CGcnJzwwQcfIDQ0FF5eXmjbti1GjhwJACgqKoKDgwMmTJiA2NhYREREYNu2bbhx4wYA4Ny5c7C3t8e2bdsQFxeH06dPo02bNpgxY0a5610TMfASEZmBmU3UYfXqDvX2hU+pt28bVPZ1ok8D+7406oNrDLzyGHiNqCKBN6egCC4Tj1dLySkof09pYGAgBEHAwYMH9R6nGXg3bNgAR0dHSY/wiRMnUKdOHaSmpiI9PR2CIODs2bM6r/Xmm29qBewdO3bAycmp3PWuiRh4iYhqqKzbYpCd4yTtnc2+oz4m4bwYes8tFnuBawAGXnkMvEZkiYH34sWLFQ6848ePx+uvvy7Zn5mZCUEQ4OvrCwAYNWoUbGxsMGjQICxbtgzJycmqY5s2bYoGDRqgUaNGqtKgQQMIgoCcnJxy172mYeAlolqn9DeKeVlAfrbuY6tDcRHgNVv3uFxdsyvUsKXhGXjlMfAaEYc0lD/wAsDVq1cxd+5c9O7dG3Z2drhw4QIAoEGDBliwYAFiYmK0SkkN+a26Mhh4iajWyL4DeM6Ujnc98qM0ULoNr/4A+fcQ3WH3yrbqrVc5MfDKY+A1Ikt9aG3AgAEVemitrCENuvTq1Qv/+584sL9Pnz748ssvDX0b1c7c2wERUbmVZxovV3sgeK94fHGROK1XcaHp6qj5cJqyxPuWfV4NwsArj4HXiCw18MbFxaFFixaqacmio6MRERGB5cuXo3PnzgCkgTcnJwdOTk748MMPERoaCm9vb7Rr10710Fp8fDwmTZqEgIAAJCYm4tSpU3j88cexZs0aAICHhwfq1auHGTNmICwsDBEREXBzc8OUKVOq5f4NxdzbARFRmYL+ATxn6Q63PvN0b/+rg/R1xg3j1S/uLLBrGHBysrgSmub75plfaGTglcfAa0SWGngBIDk5GT/88ANcXFxgbW0NZ2dnDB48GD4+PgAqNi1Zamoqhg4dCicnJ1hbW8PFxQXTp0+XDFfw8PBAnz590LBhQ9jb26Nnz57YsGGDSe/Z0CyhHRARaUkJke/BDT8MJF9TP+RVXAQEbtDf63vNTePaoUBqmGGGPuh7z6L8ql+/GjDwyqtRgdfX1xeDBg2Ck5OTzoULRo4cCUEQJOWdd96RHJOeno7hw4fjscceg4ODA7788kvJfK8AEBwcjL59+8LGxgYtW7bEggULtOqyd+9edOrUCTY2NujWrRtOnDhR4fux5MBLVcd2QEQWJzdDd4A8MQGIPiN/3r1Y+Z5gV3vgbrQ4xEFz28FvgbTIytUzwU/+vXzmV+6aNQADr7waFXjd3d0xZcoUHDx4UDbwDhgwACkpKapy//59yTEDBgzAc889h4sXL+L8+fNo3749hg0bptqflZWF5s2b47PPPkNYWBjc3NzQsGFDrF+/XnWMv78/6tati4ULFyIiIgJTp05F/fr1ERoaWqH7YeAlfdgOiKhGK8oHLq4DHt579Lqg7Dlkb/2rHSDXvVr+97wbIxNEHYAF7cpe1ay8Vr6kfZ2F7cUpxnLvl31+DcXAK69GBV5NcoF3yJAhsudERERAEARcvnxZte3kyZOwsrLC7du3AQBr1qyBo6MjCgoKVMdMnDhRtSQuAHzyyScYOHCg5Novv/wyvvnmmwrdAwMv6cN2QETVRtewgUub1PuLC+V7QBe0Ff9c00f888DXQEEOsKk/sKGfuG3F80DkCcDjD3FfZZSUiEMgNryhf/hBSkj5r7m0m/Rcjz/E4RE1aWq0KmDglWd2gdfBwQHNmjVDx44d8e233+LevXuq/Zs3b0bjxo0l5xQVFaFu3bqqeWNHjBihFZq9vb0hCIKqt7hVq1ZYunSp5Jjp06fj2WefrdA9MPCSPmwHRGQ0eVniA2N3orT3pUXKh8ftQ8Xpwco7q4Jc2TPCcPdS+uG25GvSe/DWMU3m7SDgTKnVOE9NlV5nZhPxWhaEgVeeWQVeNzc3HDlyBCEhITh06BCefvppvPTSSyguLgYA/Pnnn+jYsaPWtZo1a6Z64v/tt9/G119/LdkfHh4OQRAQEREBAKhfvz527dolOWb16tV44okn9NY5Pz8fWVlZqpKUlMTAS7LYDojIaLb9Vx3skq5I9y19puIBVnMO3bJKvK84/MFQ1vaVXr9E/DcfB79Vj+UtTXnsoe+BB6lAWoT0GreDgId3DVfHGoKBV55ZBd7S4uLiIAgCPD09AVR/4HV1ddV6qI6Bl+SwHRCRQZSUAIs6i9N5nZyk/bW9q704swEApMert63uLQ4Z0BdeNUNhwUP19ohjQPRp7eONsUCD5vU1Z0/451Ppvsxbus9RlhUviH9qDt2wMAy88sw68ALisrPr1q0DUP1DGtjDSxXBdkBElZaVLH6Vf/iH8ve8lu7lzMsSZ07QHAub8ujh7LMLAL/lZdcjNUx6TWMI2Sde+69SHVq67nHVy2V/DpEVn3XJXDDwyjPrwJuUlAQrKyscOXIEgPqhtStX1F/fnDp1SudDa4WF6tVbJk+erPXQ2qBBgyTv1bt3bz60RgbFdkBElXb4e/lAt/NjYOFTuntgNXt3AbF3+OhPwKWNla9LYS6w+zMxmBqDQgHcugLkP5BuVwbhihblkAgLxMArr0YF3uzsbAQFBSEoKAiCIGDJkiUICgrCjRs3kJ2djQkTJuDChQtISEiAp6cnnn/+eXTo0AH5+eqvOAYMGIAePXogMDAQfn5+6NChg2RasszMTDRv3hwjRoxAWFgYdu/eDVtbW61pyerVq4dFixYhMjISrq6unJaMDI7tgIgqRKEArmyVhre1/9EOdHdj1OeUflBLWYqLqu02DOpOlP5wC0inOjvyY/XW18gYeOXVqMDr4+OjcwzsyJEjkZubi/79+6NZs2aoX78+XFxcMHbsWKSmpkqukZ6ejmHDhsHOzg729vYYPXq03oUnnJ2dMX++9iTTe/fuRceOHWFtbY2uXbty4QkyOLYDIioX5api2wZpB7qs2+KUWtGnxbllA3WsQJl0BTg6TiP0/c+09Te20P26w+7SZ9THJAYANy5WXx1NhIFXXo0KvJamNgfe8o7BrqmUv3xlZGQY7T1qQzsgoiryWy7fe3lsfPmvk3VbfV5ysPHqWx0UCnEs8r3YUmOUy1gkwwIx8Mpj4DUiSw68KSkp+PHHH9G2bVtYW1ujZcuWGDRokGrGDAbesllCOyAiIwrcoDvo3gwUZ0yoqOC94uwKlqqkWP0ZlR7vW0sw8Mpj4DUiSw28CQkJePLJJ9GlSxfs378f169fR1hYGBYvXqx6+I+Bt2zm3g6IyMDCD4sLPxTmicMUdPXoxnqphziQtuw74ry7tRQDrzwGXiOy1MD77rvvwtnZGQ8favcwKANi6cAbEhKCfv36oUGDBmjSpAnGjh0rGVvt4+ODl156Cba2tnBwcECfPn2QmJio2n/48GH06NEDNjY2aNu2LWbMmIGiovI9dCEIAjZu3IihQ4eiYcOGaN++vWpmD6UTJ06gQ4cOaNCgAV5//XVs3bpVEnjv3buH//u//8OTTz6Jhg0bolu3blpzNT98+BAjRoxAo0aN0KJFCyxatAivvfYaxo0bp7Ne5t4OiMhAEgPElcnkhi6cnFTdNSQzwcArj4HXiCoUeBUK8Suq6igV6C1IT0+HlZUV5s7VsZSjBs3A+/DhQzg5OeGDDz5AaGgovLy80LZtW4wcORKAOFeyg4MDJkyYgNjYWERERGDbtm24ceMGAODcuXOwt7fHtm3bEBcXh9OnT6NNmzaYMWOG3Ntr1aVly5bYtWsXYmJi8NNPP8HOzg7p6ekAgJs3b8LGxga//PILoqKisHPnTjRv3lwSeG/duoW//voLQUFBiIuLw4oVK1C3bl0EBgaq3ue7775D69at4enpiZCQEAwaNAiPPfYYAy+RqWTdBq65qYPilveA09PE3lJAnD4rJ/3RA03dgF3DgIij1VffjJvAzUtlT6NFVE4MvPIYeI2oQoFXcwUbU5cKjAULDAyEIAiqhTzkaAbeDRs2wNHRUdIjfOLECdSpUwepqalIT0+HIAg4e/aszmu9+eabWgF7x44dcHJyKledBUHA1KlTVa8fPnwIQRBw8uRJAOI8zF26dJGcM3HixDKHNAwcOBC//vorAHFKPWtra+zdu1e1Pz09HQ0bNmTgJTIFhUL+77hFnfT/HRjlDrj/DsT5mKauJSXl//tZc4oxojIw8Mpj4DUiSwy8Fy9erHDgHT9+PF5//XXJ/szMTAiCAF9fXwDAqFGjYGNjg0GDBmHZsmVITk5WHdu0aVM0aNAAjRo1UpUGDRpAEATk5OSUWWdBECRBFADs7e3x999/AwCGDh2K0aNHS/YfPnxYEniLi4sxa9YsdOvWDY6OjmjUqBHq1auHjz/+GABw7do1CIKg6pVW6t69OwMvmSeFAki6DIQdAtyGA7Obi69rmowbwOUtwIGvDfP3ISCOoTXWOFnvudrvWXp1sMJcIPyIcZbpJYvGwCuPgdeIOKSh/IEXAK5evYq5c+eid+/esLOzw4ULFwAADRo0wIIFCxATE6NVSkpKyqyzrgfoHBwcsHXrVgDlC7zz5s3D448/jh07duDatWuIiYnBwIEDVctUM/CSxbifCOwdKR8Ia9IDU+FH5OuZcB5Y2F57u8cfQGYSMMdJ93m3g8Q/Fz4l9sRWRHo8cGkTsOIFceWyQ98BnjOBqJPiNWc01n6/wlzx3JuXgHWviuN5iSqJgVceA68RWepDawMGDKjQQ2tlDWnQpVevXvjf/8TJ0fv06YMvv/yy0vUtK/BOnjwZXbt2leyfNGmSJPAOGjRIUoeSkhJ06NBBFXizs7NRv359SU/y/fv3YWtry8BLNV/BQ+DGBSBgVfl6Qc8vrd76KhSA/0r5+q17RTyuKF97353r6uukRQCx3sCtf3VfZ14rsUe2LJc3V643OUz/N2VEFcXAK4+B14gsNfDGxcWhRYsWqmnJoqOjERERgeXLl6Nz584ApCEzJycHTk5O+PDDDxEaGgpvb2+0a9dO9dBafHw8Jk2ahICAACQmJuLUqVN4/PHHsWbNGgCAh4cH6tWrhxkzZiAsLAwRERFwc3PDlClTylXfsgLvjRs3YG1tjQkTJiAqKgr//PMPWrRoIQm848ePR6tWreDv74+IiAiMGTMG9vb2qsALAN9++y1cXFzg5eWF0NBQDB48GHZ2dgy8VP3uxYo9iHL0hbKA1eKcpqW35xpvyr4yla7Lpv7iHKy6FhpIDhaX1728BbgbLX/NY+N13/8cPc8KFBXoDtWu9rp7l/98EljVE/CcVfHeY6JyYOCVx8BrRJYaeAEgOTkZP/zwA1xcXGBtbQ1nZ2cMHjwYPj4+ACo2LVlqaiqGDh0KJycnWFtbw8XFBdOnT5cMV/Dw8ECfPn3QsGFD2Nvbo2fPntiwQccSmjqUFXgB4NixY2jfvj1sbGzwyiuvYMuWLZLAm56ejiFDhsDOzg5PPPEEpk6dii+++EISeLOzs/H555/D1tYWzZs3x8KFCzktGVW/0oFMc0jCvtH6w26Cn/rY0vtmNxdXATM1XQ+nGcLip6XXPPS9/mEcPvPV+zWHKqx4QRxOAQDHfpZek2NyycgYeOUx8BqRJQdeKh8GXqo2+h7iyrwlHqPVa3tf3J6frR5bqhGfEK0AACAASURBVJR0WX3cypcMGzbLqzAPuHFR/d7rXwfyDPQPe+nPIvmaxjCIKOD4L8BfHYHzS4D9X+n+XHWtYpYSot5fVGCYuhLJYOCVx8BrRAy8xMBL1SItQn/P7aqe4pAEzW3leVgqN0MMnbevqs+7tFF8DQBZyeK8tuX5uj49Dog+U/6v9kvfw6ym4jAGQ9Gcv3dVT93vqa94/AEUyMwak5POIQxkEgy88hh4jYiB1/h27twpma5Ms5SeW7c6MPBStdg6UBrGNr+jHdDmtVL/XFxYsetn3NC+XulxvmXRPLb40aqJJSXidUorKpAev/19IHhPxepcHsWlVm8sT9C9tMnw9SCqJAZeeQy8RsTAa3wPHjzQOV1ZTEyMZGnimojtgCot85b++bOP/iSGMfeJ4uuSEuDAWN2BbffnFX9/XeNo0+P0B96SYuDqDvUQBM1jI48DqeHq1//+LT13vkv1DAtQfo6aZYaj9HVOuunqQ1QGBl55DLxGxMBL+rAdUKX4LZMGrgtrxIULwg+rj/l7sLgvYJV6m64FD1ztgX+3V64ey57V3/MZ5S493mu2uH2+izgVWOnQvXmAdJvyQbG7MRXrOTYk5TLEmiUzqfrqQ1QGBl55DLxGVFbgzc3NlTmTaoPc3FwGXtIt+w5wP0EMfUX54sNkmbfEXlJ9IfNBinT8btIV9TXzMrWHOrjai+GzMja+pb8ux8arj5WbukuznJoifT23FbB9qHRb8F75+hiL5vsHPpoZJugf8fX+MaavD5EeDLzyGHiNSK7hFRcXIyIiAvfu3aummlFNkJmZiYiICBQWVnD8JFm2hPPyoVDZc6uvzG8j/rnxLd3X1zz24rrKP0y1upeOwJ0KeP8p/rxtkPpYzSEJVSnxvrLVMZqUUHW41XxI7l6sYR+aIzIABl55DLxGpK/hJScnq0Jvbm4u8vLyWGpRycnJUY0zVtSkpVrJ+K57ABveUM9skHRFnCEg/AgQc6b84a/0g1z6elg1HfxWnF5LOQVZZcX5aL8nIO1Fzr4jbtNVvxmO4lhdzW17Rui/J/6/QqQXA688Bl4j0tfwFAqFKvSy1M4SFRWFggLOy2mRFApx3toHpZbOLr2EbVnDAlztxSB86Dvd4fLWFWBtXyBol3q7z3xxPG+hnqEyhuqZzH8A3AwE5rYUpxgDgKPjpPWc/YT09cqXgIePvt3Ky1JvP/6ruK30fSZdEYc7ZCYZps5EFoyBVx4DrxGVp+EVFxdXe28jS/WUEs7LaXmKCsSvujUD271YcV9hbvl6bnd/Li5uoBlY1/xHd+CtibLvyN9bjKf28Vm3gStb1QtdJAaoj0+PM2nVicwdA688Bl4jYsMjqgVKzzig1UN7uOyHzVztgaiT2vPAKvmvlB6rr/e2JpC7R7mFGUoL3a8e8kFE5cbcIY+B14jY8MhiZd4SA4zvX9Vdk+pXnl7bYz9LXyt795Wvbwfpf4/iIiDWy3DL6Bqb3OfAbzWIjIq5Qx4DrxGx4ZHF0gwxWcnVXRvTe5Aizl+ra7qtO1HiMXKhT/Nr+gep4jhcS6Prvv90ru5aEVk85g55DLxGxIZHFknzQSPlV/amUpQv9obKffVfGQfGAm7Dyz8DQEkxMPPxsnswS+8741p2T66lSPBT3/f5JeLMDXdjqrtWRBaPuUMeA68RseGRRdo7ShrkQvaZ7r1Xvax+3+jTlb9Oahhw4jfAd6H6evcTyj6vpES+57Z0YFYoxJkLXO1r5wIFh78X5w3mMAYik2HukMfAa0RseGQxbgYCp6YCoQe0g17QLtPVo/R7RxwzzHVc7YEbF9X77yeotyddATJuaJ+3+R3Ac5b484nf5N+rKJ/zxxKRSTB3yGPgNSI2PLIINy7oDohr+oh//rvddHXRVY/DP4grhpV3BgCFQvd1vP8EFrQDdnxQvgfRlLLTGGiJqEZg7pDHwGtEbHhkEUoHvc3vAJ4zxXGvrvbA5c2mq8u6V+UD6MnJ5btGdpr0vCVdyxdwNcvpaca9TyKiSmDukMfAa0RseGQR5rVWB73Q/ertym3Luxu/DsWFQNA/wNxW+oPogbFlX0vz+JgzwPrXtK9z8BtxmILc+xjyoTkiIgNh7pDHwGtEbHhk1ooKpL2f55dK92sGwId3jVsXzfda0E49dvbCGt0h+Mj/xD9jvfVfCwDOLpRuW/qM9Pj8B+KfycHiEr9c4paIaijmDnkMvEbEhkdmy2u2dojMSZceo7mYwuKnxXGsCoX4wFdlx7SGHRSHShQXqrelx2sE2R/FadEUCnH+X4VCnAlAX6/vyUnS91AOi/BfIb7ePlR97IW1las3EVENwNwhj4HXiNjwyCwpe081y6kp2seVXi531zBg+/viz2cXisfcuS6uxlaeB8oSzquvtXekON1Z6XoUFWiflxIKzGqmP/SWFKuPVwbka7vF1wGrOFSBiCwCc4c8Bl4jYsMjs1L6YS5Xe2BTf+DQ9/LnlDWLga5ZDXQpzJMeq2uYgr56AGLP7/FfgLMLtM9d1Fnd67zxLel45KJ8cQhExNHyfU5ERDUUc4c8Bl4jYsOjGkVX76jm0IPSITEtouxrKo8tPQ7W1V5cgU3zdfI13dfQt5iDshz9qWL3uv8r7WtEnZSuAHbnesWuSURUwzF3yGPgNSI2PKox0iK1g2d2GjCnhTh8IOgfaTic06J811UeH+dTdmi9Gy2eo1CIPysXZPBfUfa5MWcqdr8FOeLYXUloHlf+HmciIjPE3CGvSoG3uLgYvr6+yMjIMFR9LAobHtUIuhZa2P25fLgM2Qc8SCnftY/8T1yooTC37NDq/ad4TuAG9bblPaTH3ImS9hDv+UI9HrgyNK+98U0GXiKyaMwd8qrcw2tjY4P4+HhD1MXisOFRlRXmAqt6An7LKnf+5c1lB1Fl2fCG7mm8ykvuuou7qH++fgr4e4j2MUd+BO7FVv695Rz5UWaIxDjDvxcRUTVj7pBX5cD7wgsvwNPT0xB1sThseFRlHn/o7pG8ulOcj1bfg1Z5meULuqt7i0MeqspvGbDieem1E/3Fh82Ur3WFXWP3tq7tW/UhEkREZoC5Q16VA+/JkyfRvXt3HDt2DMnJycjKypKU2owNjypNoQAiTwBznHSHQs3wduI37fN95mmHvOIi4EGqONet5nbNKbsM4fwS4E9ncbowALh9tezQbUxxZ037fkRE1YS5Q16VA6+VlZWq1KlTR1WUr2szNjyqtK0DdfRKeopBuPT8t672wN0Y6fnK7fNaAz7ztWdoCN4jrigWtMs49S8dovWF3RMTjFMHTQUPxc8u85bxV4UjIqomzB3yqhx4z549q7fUZmx4VGG6eiMlD169pXue2YTzYk/qjg+k20svB1xdEvyApd1031NlV2UjIiIJ5g55nJbMiNjwqEJuXdEOg+tflw+/s58AFj4l/hxxVPcx+dnVfVdSiQGPHlL7n9jbqrmEMBERVQlzhzwGXiNiw6MKmdFYO7DqWv3M1R7Y8i5wcb3+3uDdn1f3HemWm8FeXSIiI2DukMfAa0RseFSmBymPAux7wF8dpIE1J108RtfwBaUlMsME1vQRH1AjIqJag7lDXo0KvL6+vhg0aBCcnJwgCAIOHTok2a9QKDBt2jS0aNECDRo0wJtvvono6GjJMenp6Rg+fDgee+wxODg44Msvv0R2tvRr3eDgYPTt2xc2NjZo2bIlFixYoFWXvXv3olOnTrCxsUG3bt1w4sSJCt8PGx6p3LkOpIZrb9cVVv2WS485v0Tcfuh7cWEGTZEn1Of5/iUOiyAiolqJuUNejQq87u7umDJlCg4ePKgz8M6fPx8ODg44fPgwgoODMXjwYLRt2xZ5eXmqYwYMGIDnnnsOFy9exPnz59G+fXsMGzZMtT8rKwvNmzfHZ599hrCwMLi5uaFhw4ZYv3696hh/f3/UrVsXCxcuREREBKZOnYr69esjNDS0QvfDhkcAxOnAlKE085Z6e0GONOh6zQFivYCSEun5ytkFdA0DKC4Ue3l3DdPeR0REtQpzh7waFXg1lQ68CoUCLVq0wF9//aXalpmZCRsbG7i5uQEAIiIiIAgCLl++rDrm5MmTsLKywu3btwEAa9asgaOjIwoK1NM0TZw4EZ06dVK9/uSTTzBw4EBJfV5++WV88803FboHNrxa4H4CcPgH3VNdZdwUhxVEn1aHWrfh6v2ai0qUDrlEREQVxNwhr8qBNzU1FZ9//jmcnJxQt25dyVy8VZmHt3TgjYuLgyAICAoKkhz36quv4qeffgIAbN68GY0bN5bsLyoqQt26dXHw4EEAwIgRIzBkyBDJMd7e3hAEAffv3wcAtGrVCkuXSqdzmj59Op599tkK3QMbXi2gazGDnHTg5GT5h8nWvSr2zKpmW2heffUnIiKLwdwhr8qBd8CAAejSpQvWrFmDQ4cO4fDhw5JS6YqVCrz+/v4QBAHJycmS4z7++GN88sknAIA///wTHTt21LpWs2bNsGbNGgDA22+/ja+//lqyPzw8HIIgICIiAgBQv3597NolnZB/9erVeOKJJ/TWOT8/X7LKXFJSEhueJQveKw2yxUVij29Zq4q52otL7ip/vuZW3XdCREQWgIFXXpUDr52dnVavqyGYY+B1dXWFIAhahQ3PwtyNBorytUNs7n3g1FTt7XOcxDlz5cIvV/4iIiIDYOCVV+XA+/TTT+Pq1auGqIuEOQ5pMKseXs6DWjma43FLlwNfS+fSne8CXN0pnqdQAIEbdJ9HRERkAAy88qoceE+dOoX+/fsjISHBANVRk3tobdGiRaptWVlZOh9au3JFPTXTqVOndD60VlioXuFp8uTJWg+tDRo0SFKf3r17W85Da+cWATObAPu/qu6amA+FApjfpnzDFVztAd+Fuq+T6A/8PUR9nM88094HERFZrBqbO2qAKgfexo0bw9raGnXq1IGdnR0cHR0lpSKys7MRFBSEoKAgCIKAJUuWICgoCDdu3AAgTkvWuHFjHDlyBCEhIRgyZIjOacl69OiBwMBA+Pn5oUOHDpJpyTIzM9G8eXOMGDECYWFh2L17N2xtbbWmJatXrx4WLVqEyMhIuLq6mue0ZKXHk85sAsxtJd0W5V49dTMHD1LEcbk56dqBdv3rwOHvgbizugNvZlLZ1w7eAxQV6D+OiIionKo9d9RgVQ6827Zt01sqwsfHR+cY2JEjRwJQLzzRvHlz2NjY4M0338T169cl10hPT8ewYcNgZ2cHe3t7jB49Wu/CE87Ozpg/f75WXfbu3YuOHTvC2toaXbt2Nb+FJ4oKytcTObeV6etWkygUgPdc4Npu9TAPzenC5ErYQfU1dO3Pf1A990NERLUWA6+8GjsPryUwecOLOKo/pHnNBjxnaW8vKTZN/WqiIz9KP4u1/9H/GZ6aCiT4Sa9x57r0mDV9OEaaiIhMjoFXnkECb3FxMfbv34/Zs2dj9uzZOHjwIIqLa3GIesSkDW/fl/IhLSddR+Vuq/dv6i8Of7Bk4YfFew1Ypd52+Pvyj8ktKQEe3pO/flYykJ8tzq/LRSSIiKgaMPDKq3LgjYmJQYcOHWBra4sePXqgR48esLW1RadOnRAbG2uIOpotkza80gFt3auAz3xxrKgc34Xq41e9bPw6Vocr27R7tZVmPq79uXn8AewdpX69sD2nDSMiIrPAwCuvyoH33XffxYABA5Ceru5FvHfvHgYMGID33nuvqpc3a0ZteAoFEH9OnPYqJUQd0DxnVew6aZHqc4sLyz7enMhNIbZ3JLB5gPp1yD7x3vMfjfVOCeWUYUREZHYYeOVVOfDa2toiJCREa/u1a9fQqFGjql7erBm94emaJquiD0tpPty26W2g4KFx6mpMCoU43EBz3OzlzdLPZVN/3eF3y7vqoKvp9HRx3lwiIiIzwcArr8qB19HREf7+/lrb/fz8KjwtmaUxesMrHd4OjK36dS6sVW/Pf2Ae02YtaCvtkfVbLr2ng9+K20t/Xona7ZaIiMhcMfDKq3LgHTFiBLp27YqLFy9CoVBAoVDgwoUL6Natm2o6sdrK6A1v1zBp7+z9xMpdJ/ma9hjXvCz1660Dq2/WAYVC7HUuKREDquYDYfnZwKqe0rpf2aYdbJVKby/M034/IiIiM8XAK6/KgTcjIwODBw+GlZUVrK2tVYtQDB06FJmZmYaoo9kyesMrKgBSww0TRjWDYFEBcPAb6bZrblV/j8ooHWh3fiyOU970tu4hCrs/k77W/GwSA9TbL22snvshIiIyEgZeeQabhzcmJgZHjx7F0aNHERMTY6jLmjWzani3/lWHwYDV2kFyUWfT1ifWC/BfUf5pw3SVjBva183LAm4HmfZeiIiITMCscoeJGXzhieLiYgQFBeH+/fuGvrTZMbuGpwyKS7vpDpDVUZeqFCIiolrE7HKHCVU58I4bNw6bNm0CIIbd//znP7CyskKjRo3g4+NT1cubNbNreMqguKGf7gBZkGP6urjaAycmaG9ztQfOLwWK8oHQA+IwjM3vAMfGi0Meki6brq5EREQ1gNnlDhOqcuB1dnbG5ctiuDh06BCcnJxw/fp1TJ06FX369KlyBc2Z2TU8ZZD8ewgwp4X484zG6u3xvqapR0lx2Q+d3bxUu5dEJiIiKsXscocJVTnw2tjYICkpCQAwduxYjBs3DgAQHx+Pxx57rKqXN2tm1/B85mv3ot5PlL7OSRfHwRpaQY64gIZCASRd0Xj/BPUxym0zHzf8+xMREZk5s8sdJlTlwNu6dWucOnUKxcXFaNWqFY4fPw4ACAsLQ+PGjatcQXNmdg3vdpA6VM5pAVz3ELfrGt5wdadh3rO4EJjbUn3d1b3UP28fKj323GJg2bNc6peIiEgHs8sdJlTlwOvq6goHBwd07twZrVu3Rn5+PgBg8+bN6NWrV5UraM7MruHd0uhZzU5Tb8+8BZxdCMxtZfiHwk5N5YNnREREBmB2ucOEDDJLw/79+7FkyRLV0AYA2LZtGw4fPmyIy5sts2t4+dn6g6Yxwqi+sBt+xDDvQUREVAuYXe4woSoF3sLCQrzxxhuIjo42VH0silk2vLxMoDBX976qBt4HKYDbcCDWW71txfPq67kNF7eVlHDYAhERUQWZZe4wkSr38DZt2pSBV4bFNbzjv0gDb2ZS2edo0hwSobS8u/j60qbqW76YiIjIAlhc7jCgKgfen3/+GRMnTjREXSyOxTU8hQK4E1WxIQcRx8Se2/R43b3Df3UUXydfM169iYiIagGLyx0GVOXA++OPP8Le3h4vvPACvv76a4wfP15SajOLbXhHfxJD6pkZ+o+7tFF+fG56vHjMvEe9vneuG7/eREREFsxic4cBVDnwvv7667KlX79+hqij2bLYhuc1Wx1cI4+LY3N1KWvp3zX/AWY4ij9n3DDtPRAREVkYi80dBmCQWRpIN4tteJpz5braA64OQNhB6THpcdoBd/dnuoPvkq7iEsFERERUaRabOwyAgdeILLbhldVzm3RZe9vDe7rP3TtKXL2NiIiIqsRic4cBGCTwXr58Gb/99hs+/fRTvP/++5JSm1lswysr8OoKwEoX1kj3lRRX330QERFZEIvNHQZQ5cDr5uaG+vXrY9CgQbC2tsagQYPQsWNHODg4YNSoUYaoo9my2IZ3fmnFAm9p2WlA9h327BIRERmQxeYOA6hy4H3mmWewatUqAICdnR3i4uKgUCgwduxYTJ8+vcoVNGcW3fDSIoAYT/mQu3+M+OepqdVdUyIiolrBonNHFVU58Nra2iIhIQEA0KRJE4SEhAAAIiIi0KJFi6pe3qzVioZ3ZZt22J3vov75mlt115CIiKhWqBW5o5KqHHidnZ1VIfeZZ57Brl27AAABAQGwt6/E8rMWpFY0vMJc/cMZcu9Xdw2JiIhqhVqROyqpyoF32LBhWLx4MQBg1qxZaNasGcaMGQMXFxc+tFZbGt41t/KP3yUiIiKjqDW5oxKqHHjT09Nx+/ZtAEBJSQnmzZuH//73v/jll19w/37t7t2rVQ1PGXC9/wRmNmHgJSIiMrFalTsqiPPwGlGtanhR7oD/SvHn20HqldSIiIjIJGpV7qgggwTe2NhYTJkyBf/3f/+HtLQ0AIC7uzvCwsIMcXmzVasbXnYaUFxU3bUgIiKqNWp17ihDlQPv2bNn0bBhQ7z11luwtrZGXFwcAGDevHn48MMPq1xBc8aGR0RERKbC3CGvyoG3V69eqofWlPPwAkBgYCCcnZ2renmzxoZHREREpsLcIa/KgbdRo0aIj48HIA28CQkJsLGxqerlzRobHhEREZkKc4c8g8zD6+/vD0AaeA8ePIh27dpV9fJmjQ2PiIiITIW5Q16VA++vv/6Kvn37IiUlBY899hhiYmLg5+eHdu3aYcaMGYaoo9liwyMiIiJTYe6QV+XAW1BQgDFjxqBevXqwsrJC/fr1UadOHXz++ecoLi42RB3NFhseERERmQpzhzyDzcN78+ZNnDhxAnv27EF0dLShLmvW2PCIiIjIVJg75FU68JaUlGD+/Pno06cPXnzxRUycOBG5ubmGrJvZY8MjIiIiU2HukFfpwDtr1izUqVMH/fv3x5AhQ9CgQQOMHj3akHUze2x4REREZCrMHfIqHXjbt2+PdevWqV6fOXMG1tbWKCkpMUjFLAEbHhEREZkKc4e8Sgdea2tr3Lx5U7LNxsYGSUlJVa6UpWDDIyIiIlNh7pBX6cBbp04d3LlzR7LNzs5OtQgFseERERGR6TB3yKt04LWyssJ7772H999/X1Xq1auH/v37S7bVZmx4REREZCrMHfIqHXhHjRpVrlKbseERERGRqTB3yDPYPLym4urqCkEQJKVTp06q/Xl5efj+++/RpEkTNGrUCB988AFSU1Ml17hx4wbee+89NGzYEM2aNcOECRNQVFQkOcbHxwc9evSAtbU1nnrqKWzdurXCdWXDIyIiIlNh7pBnloG3a9euSElJUZW7d++q9n/77bdo1aoVvLy8cOXKFfTq1Qt9+vRR7S8uLka3bt3w1ltvISgoCO7u7mjatCkmT56sOiY+Ph62trb45ZdfEBERgZUrV6Ju3brw8PCoUF3Z8IiIiMhUmDvkmWXgfe6553Tuy8zMRP369bFv3z7VtsjISAiCgAsXLgAA3N3dUadOHUmv79q1a2Fvb4+CggIAwO+//46uXbtKrv3pp5/inXfeqVBd2fCIiIjIVJg75Jll4LW1tYWTkxPatm2L4cOH48aNGwAALy8vCIKAjIwMyTmtW7fGkiVLAADTpk3TCszx8fEQBAFXr14FALzyyisYN26c5JgtW7bA3t5eb93y8/ORlZWlKklJSWx4REREZBIMvPLMLvC6u7tj7969CA4OhoeHB3r37o3WrVvjwYMH+Oeff2Btba11zksvvYTff/8dADB27Fj0799fsj8nJweCIMDd3R0A0KFDB8ydO1dyzIkTJyAIgt7lk3WNL2bDIyIiIlNg4JVndoG3tIyMDNjb22PTpk3VHnjZw0tERETVhYFXntkHXgB48cUXMWnSpGof0lAaGx4RERGZCnOHPLMPvNnZ2XB0dMTy5ctVD63t379ftT8qKkrnQ2tpaWmqY9avXw97e3vk5+cDEB9a69atm+R9hg0bxofWiIiIqMZi7pBndoH3119/xdmzZ5GQkAB/f3+89dZbaNq0qWqZ42+//RatW7eGt7c3rly5gt69e6N3796q85XTkvXv3x/Xrl2Dh4cHmjVrpnNast9++w2RkZFYvXo1pyUjIiKiGo25Q57ZBd5PP/0UTk5OsLa2hrOzMz799FPExsaq9isXnnB0dIStrS3ef/99pKSkSK6RmJiId999Fw0bNkTTpk3x66+/6lx4onv37rC2tka7du248AQRERHVaMwd8swu8JoTNjwiIiIyFeYOeQy8RsSGR0RERKbC3CGPgdeI2PCIiIjIVJg75DHwGhEbHhEREZkKc4c8Bl4jYsMjIiIiU2HukMfAa0RseERERGQqzB3yGHiNiA2PiIiITIW5Qx4DrxGx4REREZGpMHfIY+A1IjY8IiIiMhXmDnkMvEbEhkdERESmwtwhj4HXiNjwiIiIyFSYO+Qx8BoRGx4RERGZCnOHPAZeI2LDIyIiIlNh7pDHwGtEbHhERERkKswd8hh4jYgNj4iIiEyFuUMeA68RseERERGRqTB3yGPgNSI2PCIiIjIV5g55DLxGxIZHREREpsLcIY+B14jY8IiIiMhUmDvkMfAaERseERERmQpzhzwGXiNiwyMiIiJTYe6Qx8BrRGx4REREZCrMHfIYeI2IDY+IiIhMhblDHgOvEbHhERERkakwd8hj4DUiNjwiIiIyFeYOeQy8RsSGR0RERKbC3CGPgdeI2PCIiIjIVJg75DHwGhEbHhEREZkKc4c8Bl4jYsMjIiIiU2HukMfAa0RseERERGQqzB3yGHiNiA2PiIiITIW5Qx4DrxGx4REREZGpMHfIY+A1IjY8IiIiMhXmDnkMvEbEhkdERESmwtwhj4HXiNjwiIiIyFSYO+Qx8BoRGx4RERGZCnOHPAZeI2LDIyIiIlNh7pDHwGtEbHhERERkKswd8hh4jYgNj4iIiEyFuUMeA68RseERERGRqTB3yGPgNSI2PCIiIjIV5g55DLxGxIZHREREpsLcIY+B14jY8IiIiMhUmDvkMfAaERseERERmQpzhzwGXiNiwyMiIiJTYe6Qx8BbhlWrVsHFxQU2Njbo2bMnAgMDy30uGx4R1VT5RcU4dPUWVnhG49e91/DVtssYuSUQ/Rb54OO1AXh62kn0meeFlV7R2HPpJg5eTYJXZCp2XkzE8eBkBN3MQEFRSXXfRqU9zC+q7iqYLYVCgfsPC1BYXPH//rkFxcjMKURaVh5upucgOvUBriSmIzIlCyUlinKdn1dYXJlq1wrMHfIYePXYvXs3rK2tsWXLFoSHh2Ps2LFo3Lgx0tLSH8oO3AAAIABJREFUynU+Gx7VFFcS7yMiOQsJdx8iIjkLvtfvIDOnsLqrRdVglXcMXCYeN3h5e8lZdJ3uAZeJx9Fm0nEExqcjOvUBbqbnIP1hARLvPcSsY+Hou8ALf56I0Hn+Kwu88eyMU/horT/mHA9HQOw9XE5Ih1/MXZwJT8XFuHu4kngf125mIOx2JkKSMhGRnIXgpAxcTkjHxnNx+HR9AH7eHYSP1wZgzN+X8b9dV/Hz7iC4HgnDHwdDMGHvNby95KzqfV+YfQZ95nnBZeJxtJ10HAtORiLoZgZuZ+TiXnZ+pUKdpvyiYoTeysSeyzdxMjQZey7dxMyj4Zh7IgLzT0Zi8enrWOUdgy1+8TjwbxKiUh4gOCkDgfHpOBx0CzsvJmLz+Xis943FKu8YLDoVhZlHwzHjaBhmHg3HzKPhmHUsHLOPhWPx6evwjEjF2et34B9zFxfi7uHg1SRs80/A9guJ+HbHFXSeehKDV57HzKPhWO4Zja1+8Vh7NhZHrt3Goau3cODfJPhEpeFw0C3VZ/T0tJNY7RODbo/+++oqo7dewsdrA/DGIh90n3lKtX3IKj8MWeVX7nbUzdUD3WeewluLz2LAsnMYuEIs7f84gQHLzuk8Z/yeIMxzj4RfzF3cTM9BWlYeMnIKkJlbiHvZ+UjJFMN13J1sxN3JRtqDPGTmFCKnoAjF5QjZqVl5mHM8HJvOx+PazQxk5RUiNSsPMWkPEHcnGzFp2Qi/nYWgm+J/t8WnorD9QiL2XL6JDb5xmH8yEl9sDsQ3269g0oFgXE99UKU2pQ9zhzwGXj169uyJH374QfW6pKQETz75JObNm1eu89nwSEmhUCAzpxBRKQ9U/1AH3czAlcT7uJSQjgtx9+AXcxd7Lol/QboF3sCeyzex/UIiNp2Px2of8R+6kVsCMXDFOQxYdg5vLzmLNxb54LWF3ui7wAu953rixTln0GmqO7rPPIVu0z3QeerJcv0j89KcM+gx6zRcJh5H3wVeeG/5Oby/2g/vLPWFy8Tj+G3fNSw9cx2rfWKw8Vwctl9IVP1jfTI0Gd5RafCKTIVnRCpOh6fi4NUkHPg3CVv94jHXPQK/7wvGXPcIbDwXh43n4rDBNw7rfWOx9mws1vjEYrVPDFZ5i2WlVzRWeEZj2ZloLD1zHUtOX8fi09ex+FQUFp2KUv3jtvZsLLb4xWP3pRs4HHQLJ0KScSosBd5RafCLuYvA+HRcSbyPkKRMRKZkIfSW+JlffvR5B8SK4Sn8dhZi0h4g9FYm/GLu4uz1O/COTMPp8FScDE3GseDbOBwkBoE9l29iV+ANzDwajp/crmLoavU/5GP+vozJB0Mw/XAoZh0Lx+SDIejm6oHnZp7CmL8vY9KBYEw9FIpJB4Ix6UAwft8XjN/2XcOQVX54e8lZjNoSiN/3icfMOR4O1yNhmHIoBJMOBGPC3mv4eXcQftx1Fd/tvIJf9lzD7/uCMX53EH745198vf0yvtp2GV9tu4Qvt17CqC2BGLklEF9sDsTnmy7i800XMXzjBQzbcEG2Dby77BxmHA3DV9suwfVIGKYeCoXLxOP4dH0APl4XYJSQbE6lzzwvfLbxIj5Y44/3lp9D/yW+6LfIB68s8EafeV6q/4een3UaL80RQ/QrC7zRYYp7tde9ukqbSfr3dZpacz6b4Rsv4ON1Aapw3vPPM+jm6oH2f5ww+Hut9Io22r81zB3yGHhlFBQUoG7dujh06JBk+xdffIHBgweX6xrGbnjekWk4EZKMEyHJOB4slmPBt3Es+DaOXruNI4/K4aBbknLoqliUoeTAv0nYf0Us+x6VvZdvYo+yXBLL7ks3sPvSDbgFimVX4A38c1EsOy8mYufFROy4IJbtFxKxPSAB2wMS8Pejss1fLFv94rHVLx5b/OKx+bxYNp2P1wpD686KgUguFCmD0XKNcKQrIP3lEYWFHpEYvycIo7dewuebLuL/1l/Ay396ou8CL7yxyAcDV5zDkFV++HhtgFjWieWTR+XT9eL2t5ecxcfrAvDRWn+8v9oPg1f5YdCK83h32Tm8s9QXby0+i36LfPDW4rN4+U9PPDfzFDpNddf7Fz8LS3WUk6HJiL/7EApF2T1cujzML8KBf5NwOOgW5p6IwKmwFBy9dhtTDoWo3qOtRrvX/Nn1SBi+2yn2du24kIiN5+LgHZmGY8G3sfTMddVxXaZJf2F741HA7D3XEy/NEQOJy8Tj6D3XE68u9MZTk8VwMmDZOXyxORDbAxKwxicWc09EYKFHJJadicZqnxhs9YvHpQTxa/Tw21n498Z9jNgcKHmvDn8YNox1+MMdfRd4YcTmQIxzu4r/zPdS/WLx275r+GRdAF6acwbt/ziBNpOOo99fPuj3lw8+XOOPH3ddxfjdQfh9XzCmHArB7GPhWHAyEgtORmL+yUjMc4/Ex2sD8MEafwxeKf599PaSs3j90TVcJh7HszNOqe5p0oFg/LLnGibuD8aHa/zxxiIf9F3ghf+uPI8+88RfeD9Y44+P1wZg6Go/jNoSiNFbL+EZVw8MXHEOJ0KSkVdYjIycAriHJMMtUPy3Yf+VJPjH3IV/7F2cjxZ75T3CUlS/PJ69fgdpD/JQWFyi1e7yi4pxMz0HVxLvY71vLM5F34Hvo18+z4Snwj0kGZ4RqQi9lYl72flIvPcQ+64kYZ57pOS/na5f8p+afAKdp55Ep6nuqjZS0fLinDM6tz/j6oFnXD1U3xT0+8sHry70xsAV5zBqSyC+/+dfjHj0i+ekA8H4YI2/UYdkMPDKY+CVcfv2bQiCgICAAMn23377DT179tR5Tn5+PrKyslQlKSnJqA2v11zPav9Hk6XipdNUd/Sa66nqAXr9Lx+8scgH/Zf4qr6y+3CNP0ZtEb8C+9+uq5iw9xqmHArBpvPx8I5Mw7lo8SvLi3HiV75Xb9xHcFIGAmLv4d8b9xGd+gDxdx/iZnoOkjNzkZVXKBkfp1AosP1CIqYdDoVb4A2EJGUiOCkDlxLS4R97Fz5RaTgZmoJJB0Iw5u/LmHIoBJMPhuDXvdfwk9tVfLP9iqq3sP8SXwxccQ7/XXle9QvA4FV++HzTRXy7Q6z/qC2BGL8nCD+5XcVPbuJXzON3B2H8niD8sucaft17Db/tE3stJ+4PftQLGoI/DoZgyqEQTD0UimmHQzH9Ufly6yWM3x2E7//5F19uvYRhGy7go7X+GLraDwNXPOp9+8sHfeZ5ocu0k3h+1mn0/FP8B+nVhd54Y5EP3lwsfoX+wmwxOL0w+zTeXHwW7y4T72Xoaj98tNYfn6wLwPCNFzBis/iP/lfbLuOb7VfQZ54X1vvGYtKBELy5+CwWn76OZWeisehUFOa5R2La4VC4HgnDx+sCMOlACJZ7RmPxo1/Alp2JxirvGKzxicWc4+HoNdcTrkfCsMo7BotPRWGuewTmnojA0jPXsdIrGmt8YrHBNw5b/OLxd0CC6he/Db5x2OoXj50XE7Hr0S+iuy/9f3t3HhTFmfcB3CYiIetFbUQiViSrpUEhgCLocriJBpTNgqVJUFAxiqzxhOStrDHJipaVZaOu92tEI4rHazySGLMeaATMoiQQFQ9UjEQOZfBERBGB+b5/TKZnHgFFl5me6fl+qp7aZY7UM1/7+HX300/rrg5syynG9lzdAe1Xx0rw9THdAe/eU1dQcc+0w1m0Wi1u3a2R/26swLEG92vr8HnGLxjxv1lIzryIr4+VYs/JK/j+rEZe/3Iv6da905crcF5TifMa3VWcvBLdFZysX67hXg3HnJqa8fKl1Wpxv7YONbWNL3darRYP6upxt6YWu05cxprDF/HN8VJ8l6e7SnTgjG5YyMWrd3Cl4p6wvpwrqxSWbUvDgrdpLHib8DQF75w5c9CqVasGzVQL3uSNuU2ejYxcfQSjVusuYY5OPoqoNboWvSYb0Wuy5cucY9ZmY+wXP2LsF7rLn+O+0F0KjVn3o3xU/06K7jLphJSfMHG9vukuocZuyMGk31pcqq79NTUXf03NxeSNuXh3k65N2fSzrm3Wtam/tWlbjmH6b23G/x3DzN9aYwXR//zWGiuMPvxK1xorkObsOo05u05jwb5z2PpTEXb+XCKf/d6UfQlHfrmOQ2fLsf90WYMz5sZnzfUFw3d5V7Dn5BXs+23DeOhsOTLOX8UPBdfkcXMZ56/iVGkFLpTrxjBerbzPGy2IiMikWPA2jQVvE55mSIO5z/ASERER6bHgbRoL3kfw8/PDtGnT5L/r6+vh6urKm9aIiIjI4rDuaBoL3kfYunUrHBwcsH79euTn5yMuLg4dO3aERqNp1ve54BEREZG5sO5oGgvex1i+fDlefPFFtGnTBn5+fsjOzm72d7ngERERkbmw7mgaC14T4oJHRERE5sK6o2kseE2ICx4RERGZC+uOprHgNSEueERERGQurDuaxoLXhCoqKtCqVSuUlJQI05WxsbGxsbGxsbV000+HWlFRoXQJZHFY8JqQfsFjY2NjY2NjYzNXKykpUboEsjgseE2ovr4e+fn58sKn9JGfJTT9QQDzYBbMhHkwB+bBPFq2VVRUoKSkBPX19UqXQBaHBa+J3b7N8TTGmIcBs2iImYiYhw5zEDEPEfOg5mDBa2JcEUXMw4BZNMRMRMxDhzmImIeIeVBzsOA1Ma6IIuZhwCwaYiYi5qHDHETMQ8Q8qDlY8JrY/fv3MWfOHNy/f1/prlgE5mHALBpiJiLmocMcRMxDxDyoOVjwEhEREZGqseAlIiIiIlVjwUtEREREqsaCl4iIiIhUjQUvEREREakaC14iE7pz547SXbAYWq1W6S4QWY2rV69ynSFqQSx4/wv6R/fV1dUp3BPLUFxcjL179yrdDYtw6dIlhISEYMqUKQBg8495vHbtGsrLy+V1xdZ35LW1tQC4XABASUkJNm3ahJs3byrdFYvw66+/YtiwYRgxYgQALiPl5eUoLCxEVVUVAG476Omx4H1KCQkJGD16tNLdsBgFBQWQJAmOjo4oKChQujuK0Wq1iIuLQ+vWrdG+fXv06dPH5ndYU6ZMgYuLC7y8vDB48GBcuHBB6S4pasaMGQgJCVG6GxahuLgYzz//PCRJQlpamtLdUZTxtsPV1RWdOnWy+Xllp02bBicnJ3h4eMDd3R1ZWVl48OCB0t0iK8WC9wkdO3YMQ4YMQadOnWBnZ4d9+/YB4FneM2fOYOjQoejatSvGjRundHcUsWjRIrRv3x5+fn44efIkVqxYAW9vb1y5ckXprinm/fffh4+PDzIyMpCamoqgoCB4enri8OHDSnfN7PLz8xEWFoZu3bpBkiRs2rQJgG2fwdNoNIiIiEC3bt0QGhqK8vJypbukiIULF8rbjuPHj2P37t3o3bs3cnNzle6aYj799FN4e3sjMzMThw4dwvDhw/Hyyy9j48aNSneNrBQL3ie0evVqxMTE4N///jfGjBkDDw8P+T1bvtSyZcsWhIeH4+jRo7Czs0N6errSXTKrgoICBAcHIyUlRX4tMzMTkiShpKQEgG0tH1qtFnfv3kX//v2RmJgov37v3j34+PggKioKv/zyi4I9NL+dO3di4sSJOHToEOLj4+Hi4mLzZ6v27t2LiIgIFBUVQZIkpKSkyGc1bWV9qaqqQmhoKNatWye/lpeXh9atWyMnJweAbR4Uvfrqq5g8ebL8d319PUaOHIkhQ4bY9IEAPT0WvE9Io9Hg5MmTAID09HS88MIL+Ne//gXA9s7yGm+Et2/fjunTpwMAXn/9dQQHBwOAPO5K7WpqaoQdtFarRV5eHrp3747U1FQFe6ac0tJSuLi44NtvvwWgywgAtm3bBk9PTyxdulTJ7pncw0XK9evXkZ+fD0A3TrNLly6YNWtWo59VG+PfZ/z/jxw5gpEjRwIAoqOj4e7ujurqatXf7Pnwv7fxtqO+vh43btzAyy+/jKSkJHN3zSLcuHED3t7eWLZsGQDDvvXQoUMIDAxEfHy8kt0jK8WC9xE+/fRTxMfH4/PPP5d31sZu3bqFWbNmoXPnzqisrASg7h3Xo/L45JNPEBMTA0B3d7G9vT2GDh2Kfv364fjx4wr01rSaysL43//KlSvo2rUr1qxZA0DdZ6x27tyJ27dvy3/rf+uAAQMQFRUFwHCjFgBERERg+PDhuHr1qnk7aiZz587F+PHjMW/ePFy/fr3B+3V1dVi+fDnatGmDoqIiAOpdPh6VxdKlSxEWFib/7ejoiICAALi6uuI///mPubtqFk3lYXzC5MaNG+jbty8+/vhjJbpoVlu2bMGZM2fkv/XrQXh4OIYMGQJA3K7Gx8dj8ODBOH36tHk7SlaPBW8jzp07h969e8PT0xORkZFwcnLCn/70J2RnZwMQd0zHjx+Hh4cH4uLiAKiz4H1cHgDwzjvv4KuvvgIAbN68GW3btsUzzzyjurObzckCMCwHgYGB8oGAGgua9PR09OrVC5IkYfXq1fLr+t/6xRdfwN7eXr6Rsbq6GgCQlpaGZ599FqWlpebvtAkVFxejb9++8PT0xNSpU+Hi4gJfX19s374dgLgMXLt2Db6+vhg+fLhS3TWpx2UBAPPmzcPy5csBAN9++y3atWuHZ555BgsWLFCq2ybT3GVDv+0IDw+XDwbUuO3IzMyEh4cHJEnC7Nmzce/ePQCGwj89PR2SJCEzMxOA4YA5NzcXbdu2lYd7EDUXC95GLFq0CAMHDpRXsLKyMnh5eeHtt9+Wxx3q37t//z5WrFiBdu3ayUepGRkZqppi51F5nDt3DgAwadIkREVFISgoCE5OTkhMTISzs7MwflMNmrNs6HdYNTU1mDBhAsLCwlR5iTY/Px+RkZGYOnUq4uLi8OKLLza4Qa+wsBCBgYHymRq9goICdOjQAd9//705u2xy69evh7e3NyoqKgDohvSEh4cjMDAQJ06cACCe6d69e7ewU9+/fz/Onz9v/o6bwKOy+PnnnwHozta9+uqrCA4OhpOTE5YsWQI3NzfExsbK31OL5iwbxmd5582bB29vb1y7dk2R/ppSSUkJJkyYgPj4eMyePRtOTk44evSo8Jm7d+8iIiICPj4+cjEM6HJr27atcOBE1BwseB9SW1uLCRMmICIiQjiq3rZtGwYMGIAPP/xQfk3/fmFhIYYNGwZvb28EBATgueeekwtBa/e4PD766CMAuvF3HTp0QGxsLC5dugQAWLlyJSRJwtmzZxXpe0t7kmVDX/R+8MEH8PHxAaC+szQajQZr165Ffn4+Kisr4erqivfff1/4jFarxZ49e+Dg4IDPPvtMHsKQnJwMX19feSiQWiQmJqJ///5CUZuZmYnBgwcL0xjql4V79+5h1KhRcHNzg7+/PxwdHfHjjz+avd+m0Jws5s6di+effx6xsbHyAaP+IGD//v2K9NtUnnTZWLx4MXr16gWNRqO6bcft27exY8cOnDp1CgDQs2dPjBo1Crdu3RI+d+7cOXTs2BEzZ86U9yNbt26Fl5eXTc9+Q0+HBW8joqOjERISgrq6OuGIe+rUqXjttddw7NgxAIaipqCgAP7+/pAkCRMnTlTdTvxReQQHB+PXX3/F+fPnceLEiQY3X3z22WeNjn+2Vs1dNvQ7tf3798POzk61MxIYD+FZt24dHBwc5LNVxtasWYPOnTvD3d0db775JhwcHDB//nxotVpV7cxnzZqFgQMH4vLly8LrCxcuhLe3Nw4cOADAkNuFCxfw+uuvQ5IkxMbGqmrb8agsPD09kZ2djfLycpw5c6bBMrB8+XLVzUHb3GVDv+04e/YsJElCXl6e2ftqbgcOHIAkSfjmm2/kZUH/v7t27YKHhwc6d+6M8PBwtGnTBrNnz1bl8EEyLRa8RozHDtnZ2ck3W+k3QBkZGejRowe2bdsmfycnJwc9e/aEt7e3MPBeDZqTxx/+8AebuLT0NMsGAOzYsQMTJ07E9evXVVXYGTP+Xf7+/ggPDxfOYullZWVh2bJliI+Pb7Qotmb6na++SPn666+F90+cOAF/f3/hrvtz586hf//+6NOnj6puwGlOFn5+fvjnP//Z4LtqnOnmaZYNADh69Cji4uJQVlam2m0HYMhn2LBh8Pf3l6dxNFZQUIDU1FTMmTPHJg4AyDRsruA1Hgv0MP1Ourq6GoMGDZLHHRpvbLp374558+bJf1+/ft2q7yZuyTysfaPcklmo4RG6zclDT/87Dx8+DDs7O3kqsrq6OtXNxNDYv6lxHm+99RZ8fHwajL309/eXp+4DgMrKSqsv/FsiC2teRx7WEnmo6czl4/Iw/ruoqAh2dnZYtmyZnEFjxS/R07KZgvfBgweYPHkyhg4dirFjx+Lo0aPyymh8yb2urg4ajQYZGRmwt7fHqlWr5JXv5s2beOWVV7BixQoA1r2hNkUe1opZiJqbR21tLTQaTYPvR0dHo1+/fjh48CBCQ0Px8ccfW/UDFh48eIAFCxbIs5AYMz4jWVNTgwsXLqCoqAiOjo6YPXu2fINSbW0tgoOD8fe//91s/TYFZiFiHqLm5lFbWyvPSW383nvvvYeXXnoJW7ZsQUhICMaMGSPP7EL037KJgresrAw+Pj744x//iJUrV8LLywteXl4NLiEtXboUbdq0wfr16wEA8+fPh7OzM2JjY3H48GEkJCTgpZdesvqbsJiHAbMQPUkeDg4OSElJaXDgd+TIEUiSBEmSEBoaatUzluzZswfu7u6QJAnR0dHy+MuHf/PSpUvx3HPPyZfpk5OT0aNHD4SGhmLXrl1ISEjACy+8gJ9++snsv6GlMAsR8xA9aR4LFy6UD4T1nykuLpa3HX/+859x48YN8/4IUjWbKHh37NiBPn36yHN+VlRUIDExEc8++6w8di4yMhJdunTBhg0bhBV02bJlCAoKgqenJ7y8vFRxBzXzMGAWoifJIzU1Vcijrq4OGzZsgL29Pfz9/eUb+KxVVVUVYmNjMWPGDPzjH/+Ar68vVq1aJXympqYGkydPhrOzMzZu3Chcjt69ezfCwsIwcOBA+Pr6Npir2ZowCxHzED1NHg8Xwps3b0br1q3Rv39/VT6siJSn6oJXv4FZtWoVunTpIrxXVlaGwYMHy4/Azc7OFp4U9fCjMAsLC83QY9NiHgbMQvTf5KF39+5dLFmyRHgAhTXTarXIysqSpxgcOXIk/vKXvwg3zWi1WhQUFDS5fABodNiHtWEWIuYheto89Orr67Fz507VbDvIMqmu4N2+fTsOHDggzNGXnJyMvn374vDhw8JnDx48CHt7e3m+RzXdLKDHPAyYhYh5iBrLw1haWhp8fHyQmJho1eP3m4NZiJiHiHmQNVJNwZuamgpnZ2f4+fmhU6dOCAgIwI4dOwAAx44dQ+/evZGUlCTcdKPRaBAeHo6xY8cq1W2TYR4GzELEPESN5aG/6aa+vl7YYU+ZMgWDBg3CwYMHAVj3jauNYRYi5iFiHmTNrL7gra2txZIlS+Du7o61a9eipqYGWVlZGDduHIYNGyZPrRQXFwc/Pz+kp6cL3x85ciTGjx+vQM9Ng3kYMAsR8xA9Lg/jBx8Yz6Wqn0KqqqoK9fX18qOArXkOWWYhYh4i5kFqYPUFb0VFBT766CMkJSUJl1mTkpIQEBAgT/2iv/t89OjR8g05ABAWFoaEhASz99tUmIcBsxAxD9Hj8qiqqhI+rz9DlZSUhIEDB+Jvf/sb+vbtC19fX6vfgTMLEfMQMQ9SA6sseAsKCoTLI8ePH5dXIv3KuHnzZnh7ewuXZbdv346goCB069YNixYtwtixY+Hs7IwffvjBvD+ghTEPA2YhYh6ip83D+P2cnBzY29tDkiTExcVZ7aOzmYWIeYiYB6mNVRW8X375Jdzc3NCrVy/4+flh7dq1wvvGR55RUVHy5Vfjlay0tBRxcXEYPnw4wsLC5LtKrRHzMGAWIuYheto8Hn4q1KpVqyBJEkJCQnDx4kXTd9wEmIWIeYiYB6mV1RS8aWlpcHNzw8qVK7Fv3z689957sLe3R3JysvwkFq1WC61Wi+rqarzyyivYuHFjk/89a396C/MwYBYi5iFqyTzy8vLw5ZdfmrP7LYpZiJiHiHmQmll8wau/pDJ37lz069dPeETplClT4Ovr2+AxhpcvX4abmxsKCgoA6C7NqGXsIfMwYBYi5iFiHgbMQsQ8RMyDbIHFF7x6kZGRePvttwFAXhlv3ryJwMBAxMTEoKysTP7shg0bEBQUhDt37mDGjBlo3bo1RowYgQcPHqhmahTmYcAsRMxDxDwMmIWIeYiYB6mZxRW8aWlpmD59OhYvXiw8qjU5ORnt2rWTB83rV8bk5GT07NlTnkJJq9XirbfegpOTE37/+9+jT58+yMnJMfvvaCnMw4BZiJiHiHkYMAsR8xAxD7JFFlPwXrlyBW+88QacnZ0RHR0NT09PdOjQQV4Zz58/D1dXV3zyyScAxJtrXFxcsHjxYgC6x5u+8cYb6Nq1K7Zu3Wr+H9JCmIcBsxAxDxHzMGAWIuYhYh5kyyyi4L179y5iYmIQGRmJwsJC+XU/Pz/5DtDKykrMnz8fjo6OKC4uBmAYdzRo0CDExsbK38vNzTVj71se8zBgFiLmIWIeBsxCxDxEzINsnUUUvIDu6U579+4FYJjeJDExEf7+/vIKV1hYiICAAAwYMACXLl0CABQVFcHd3R3fffedMh03EeZhwCxEzEPEPAyYhYh5iJgH2TKLKXiN7wrVz/MXFRWFSZMmCZ8rLS1Fjx494ObmhjfffBNdunTBa6+9Bo1GY9b+mhrzMGAWIuYhYh4GzELEPETMg2yZxRS8jQkICMD69esB6FZO/Qp64cIFbN26FQkJCfL7toB5GDALEfMQMQ8DZiFiHiLmQbbCYgveixcvonPnzsI4IVt+LCHzMGAWIuYhYh4GzELEPETMg2yJxRW8+nFEGzZsQPfu3eXXExMTMXnyZJSXlyvVNUUwDwNmIWIeIuZhwCxEzEPEPMgWWVzBqzc5Y6JpAAABeUlEQVR16lR88MEH8qMOnZ2dsX//fqW7pRjmYcAsRMxDxDwMmIWIeYiYB9kSiyx4q6ur0aNHD0iSBAcHByQlJSndJUUxDwNmIWIeIuZhwCxEzEPEPMjWWGTBCwBDhgzBu+++i+rqaqW7YhGYhwGzEDEPEfMwYBYi5iFiHmRLLLbg1T/akHSYhwGzEDEPEfMwYBYi5iFiHmRLLLbgJSIiIiJqCSx4iYiIiEjVWPASERERkaqx4CUiIiIiVWPBS0RERESqxoKXiIiIiFSNBS8RERERqRoLXiIiBcTExCAiIkLpbhAR2QQWvERELaxVq1aPbHPmzEFFRQVu3bqldFeJiGwCC14iohZWVlYmtyVLlqB9+/bCa3fu3FG6i0RENoUFLxGRCaWkpKBDhw4NXn94SMOgQYMwbdo0zJw5Ex07doSzszOSk5NRVVWF8ePHo23btujevTv27Nkj/HdOnTqFoUOH4ne/+x2cnZ0xZswYXLt2zeS/i4jImvw/LNr18T321WcAAAAASUVORK5CYII=\" width=\"700\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f,ax=plt.subplots(figsize=(7,3))\n",
    "\n",
    "combined_df[['Close', 'Close_ndaq']].rolling(window=30,center=True).median().plot(ax=ax)\n",
    "\n",
    "ax.set(xlabel='Time',ylabel='Pearson r')\n",
    "\n",
    "ax.set(title=f\"Overall Pearson r = {np.round(overall_pearson_r,2)}\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "id": "automotive-aurora",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "/* global mpl */\n",
       "window.mpl = {};\n",
       "\n",
       "mpl.get_websocket_type = function () {\n",
       "    if (typeof WebSocket !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof MozWebSocket !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert(\n",
       "            'Your browser does not have WebSocket support. ' +\n",
       "                'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "                'Firefox 4 and 5 are also supported but you ' +\n",
       "                'have to enable WebSockets in about:config.'\n",
       "        );\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure = function (figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = this.ws.binaryType !== undefined;\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById('mpl-warnings');\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent =\n",
       "                'This browser does not support binary websocket messages. ' +\n",
       "                'Performance may be slow.';\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = document.createElement('div');\n",
       "    this.root.setAttribute('style', 'display: inline-block');\n",
       "    this._root_extra_style(this.root);\n",
       "\n",
       "    parent_element.appendChild(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen = function () {\n",
       "        fig.send_message('supports_binary', { value: fig.supports_binary });\n",
       "        fig.send_message('send_image_mode', {});\n",
       "        if (fig.ratio !== 1) {\n",
       "            fig.send_message('set_dpi_ratio', { dpi_ratio: fig.ratio });\n",
       "        }\n",
       "        fig.send_message('refresh', {});\n",
       "    };\n",
       "\n",
       "    this.imageObj.onload = function () {\n",
       "        if (fig.image_mode === 'full') {\n",
       "            // Full images could contain transparency (where diff images\n",
       "            // almost always do), so we need to clear the canvas so that\n",
       "            // there is no ghosting.\n",
       "            fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "        }\n",
       "        fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "    };\n",
       "\n",
       "    this.imageObj.onunload = function () {\n",
       "        fig.ws.close();\n",
       "    };\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_header = function () {\n",
       "    var titlebar = document.createElement('div');\n",
       "    titlebar.classList =\n",
       "        'ui-dialog-titlebar ui-widget-header ui-corner-all ui-helper-clearfix';\n",
       "    var titletext = document.createElement('div');\n",
       "    titletext.classList = 'ui-dialog-title';\n",
       "    titletext.setAttribute(\n",
       "        'style',\n",
       "        'width: 100%; text-align: center; padding: 3px;'\n",
       "    );\n",
       "    titlebar.appendChild(titletext);\n",
       "    this.root.appendChild(titlebar);\n",
       "    this.header = titletext;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = (this.canvas_div = document.createElement('div'));\n",
       "    canvas_div.setAttribute(\n",
       "        'style',\n",
       "        'border: 1px solid #ddd;' +\n",
       "            'box-sizing: content-box;' +\n",
       "            'clear: both;' +\n",
       "            'min-height: 1px;' +\n",
       "            'min-width: 1px;' +\n",
       "            'outline: 0;' +\n",
       "            'overflow: hidden;' +\n",
       "            'position: relative;' +\n",
       "            'resize: both;'\n",
       "    );\n",
       "\n",
       "    function on_keyboard_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.key_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    canvas_div.addEventListener(\n",
       "        'keydown',\n",
       "        on_keyboard_event_closure('key_press')\n",
       "    );\n",
       "    canvas_div.addEventListener(\n",
       "        'keyup',\n",
       "        on_keyboard_event_closure('key_release')\n",
       "    );\n",
       "\n",
       "    this._canvas_extra_style(canvas_div);\n",
       "    this.root.appendChild(canvas_div);\n",
       "\n",
       "    var canvas = (this.canvas = document.createElement('canvas'));\n",
       "    canvas.classList.add('mpl-canvas');\n",
       "    canvas.setAttribute('style', 'box-sizing: content-box;');\n",
       "\n",
       "    this.context = canvas.getContext('2d');\n",
       "\n",
       "    var backingStore =\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        this.context.webkitBackingStorePixelRatio ||\n",
       "        this.context.mozBackingStorePixelRatio ||\n",
       "        this.context.msBackingStorePixelRatio ||\n",
       "        this.context.oBackingStorePixelRatio ||\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        1;\n",
       "\n",
       "    this.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband_canvas = (this.rubberband_canvas = document.createElement(\n",
       "        'canvas'\n",
       "    ));\n",
       "    rubberband_canvas.setAttribute(\n",
       "        'style',\n",
       "        'box-sizing: content-box; position: absolute; left: 0; top: 0; z-index: 1;'\n",
       "    );\n",
       "\n",
       "    // Apply a ponyfill if ResizeObserver is not implemented by browser.\n",
       "    if (this.ResizeObserver === undefined) {\n",
       "        if (window.ResizeObserver !== undefined) {\n",
       "            this.ResizeObserver = window.ResizeObserver;\n",
       "        } else {\n",
       "            var obs = _JSXTOOLS_RESIZE_OBSERVER({});\n",
       "            this.ResizeObserver = obs.ResizeObserver;\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.resizeObserverInstance = new this.ResizeObserver(function (entries) {\n",
       "        var nentries = entries.length;\n",
       "        for (var i = 0; i < nentries; i++) {\n",
       "            var entry = entries[i];\n",
       "            var width, height;\n",
       "            if (entry.contentBoxSize) {\n",
       "                if (entry.contentBoxSize instanceof Array) {\n",
       "                    // Chrome 84 implements new version of spec.\n",
       "                    width = entry.contentBoxSize[0].inlineSize;\n",
       "                    height = entry.contentBoxSize[0].blockSize;\n",
       "                } else {\n",
       "                    // Firefox implements old version of spec.\n",
       "                    width = entry.contentBoxSize.inlineSize;\n",
       "                    height = entry.contentBoxSize.blockSize;\n",
       "                }\n",
       "            } else {\n",
       "                // Chrome <84 implements even older version of spec.\n",
       "                width = entry.contentRect.width;\n",
       "                height = entry.contentRect.height;\n",
       "            }\n",
       "\n",
       "            // Keep the size of the canvas and rubber band canvas in sync with\n",
       "            // the canvas container.\n",
       "            if (entry.devicePixelContentBoxSize) {\n",
       "                // Chrome 84 implements new version of spec.\n",
       "                canvas.setAttribute(\n",
       "                    'width',\n",
       "                    entry.devicePixelContentBoxSize[0].inlineSize\n",
       "                );\n",
       "                canvas.setAttribute(\n",
       "                    'height',\n",
       "                    entry.devicePixelContentBoxSize[0].blockSize\n",
       "                );\n",
       "            } else {\n",
       "                canvas.setAttribute('width', width * fig.ratio);\n",
       "                canvas.setAttribute('height', height * fig.ratio);\n",
       "            }\n",
       "            canvas.setAttribute(\n",
       "                'style',\n",
       "                'width: ' + width + 'px; height: ' + height + 'px;'\n",
       "            );\n",
       "\n",
       "            rubberband_canvas.setAttribute('width', width);\n",
       "            rubberband_canvas.setAttribute('height', height);\n",
       "\n",
       "            // And update the size in Python. We ignore the initial 0/0 size\n",
       "            // that occurs as the element is placed into the DOM, which should\n",
       "            // otherwise not happen due to the minimum size styling.\n",
       "            if (fig.ws.readyState == 1 && width != 0 && height != 0) {\n",
       "                fig.request_resize(width, height);\n",
       "            }\n",
       "        }\n",
       "    });\n",
       "    this.resizeObserverInstance.observe(canvas_div);\n",
       "\n",
       "    function on_mouse_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.mouse_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousedown',\n",
       "        on_mouse_event_closure('button_press')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseup',\n",
       "        on_mouse_event_closure('button_release')\n",
       "    );\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousemove',\n",
       "        on_mouse_event_closure('motion_notify')\n",
       "    );\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseenter',\n",
       "        on_mouse_event_closure('figure_enter')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseleave',\n",
       "        on_mouse_event_closure('figure_leave')\n",
       "    );\n",
       "\n",
       "    canvas_div.addEventListener('wheel', function (event) {\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        on_mouse_event_closure('scroll')(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.appendChild(canvas);\n",
       "    canvas_div.appendChild(rubberband_canvas);\n",
       "\n",
       "    this.rubberband_context = rubberband_canvas.getContext('2d');\n",
       "    this.rubberband_context.strokeStyle = '#000000';\n",
       "\n",
       "    this._resize_canvas = function (width, height, forward) {\n",
       "        if (forward) {\n",
       "            canvas_div.style.width = width + 'px';\n",
       "            canvas_div.style.height = height + 'px';\n",
       "        }\n",
       "    };\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    this.rubberband_canvas.addEventListener('contextmenu', function (_e) {\n",
       "        event.preventDefault();\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus() {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'mpl-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'mpl-button-group';\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'mpl-button-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        var button = (fig.buttons[name] = document.createElement('button'));\n",
       "        button.classList = 'mpl-widget';\n",
       "        button.setAttribute('role', 'button');\n",
       "        button.setAttribute('aria-disabled', 'false');\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "\n",
       "        var icon_img = document.createElement('img');\n",
       "        icon_img.src = '_images/' + image + '.png';\n",
       "        icon_img.srcset = '_images/' + image + '_large.png 2x';\n",
       "        icon_img.alt = tooltip;\n",
       "        button.appendChild(icon_img);\n",
       "\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    var fmt_picker = document.createElement('select');\n",
       "    fmt_picker.classList = 'mpl-widget';\n",
       "    toolbar.appendChild(fmt_picker);\n",
       "    this.format_dropdown = fmt_picker;\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = document.createElement('option');\n",
       "        option.selected = fmt === mpl.default_extension;\n",
       "        option.innerHTML = fmt;\n",
       "        fmt_picker.appendChild(option);\n",
       "    }\n",
       "\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.request_resize = function (x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', { width: x_pixels, height: y_pixels });\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_message = function (type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function () {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({ type: 'draw', figure_id: this.id }));\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function (fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] !== fig.canvas.width || size[1] !== fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1], msg['forward']);\n",
       "        fig.send_message('refresh', {});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function (fig, msg) {\n",
       "    var x0 = msg['x0'] / fig.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / fig.ratio;\n",
       "    var x1 = msg['x1'] / fig.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / fig.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0,\n",
       "        0,\n",
       "        fig.canvas.width / fig.ratio,\n",
       "        fig.canvas.height / fig.ratio\n",
       "    );\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function (fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function (fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch (cursor) {\n",
       "        case 0:\n",
       "            cursor = 'pointer';\n",
       "            break;\n",
       "        case 1:\n",
       "            cursor = 'default';\n",
       "            break;\n",
       "        case 2:\n",
       "            cursor = 'crosshair';\n",
       "            break;\n",
       "        case 3:\n",
       "            cursor = 'move';\n",
       "            break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_message = function (fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function (fig, _msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function (fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_history_buttons = function (fig, msg) {\n",
       "    for (var key in msg) {\n",
       "        if (!(key in fig.buttons)) {\n",
       "            continue;\n",
       "        }\n",
       "        fig.buttons[key].disabled = !msg[key];\n",
       "        fig.buttons[key].setAttribute('aria-disabled', !msg[key]);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_navigate_mode = function (fig, msg) {\n",
       "    if (msg['mode'] === 'PAN') {\n",
       "        fig.buttons['Pan'].classList.add('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    } else if (msg['mode'] === 'ZOOM') {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.add('active');\n",
       "    } else {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message('ack', {});\n",
       "};\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function (fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = 'image/png';\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src\n",
       "                );\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data\n",
       "            );\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        } else if (\n",
       "            typeof evt.data === 'string' &&\n",
       "            evt.data.slice(0, 21) === 'data:image/png;base64'\n",
       "        ) {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig['handle_' + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\n",
       "                \"No handler for the '\" + msg_type + \"' message type: \",\n",
       "                msg\n",
       "            );\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\n",
       "                    \"Exception inside the 'handler_\" + msg_type + \"' callback:\",\n",
       "                    e,\n",
       "                    e.stack,\n",
       "                    msg\n",
       "                );\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "};\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function (e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e) {\n",
       "        e = window.event;\n",
       "    }\n",
       "    if (e.target) {\n",
       "        targ = e.target;\n",
       "    } else if (e.srcElement) {\n",
       "        targ = e.srcElement;\n",
       "    }\n",
       "    if (targ.nodeType === 3) {\n",
       "        // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "    }\n",
       "\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    var boundingRect = targ.getBoundingClientRect();\n",
       "    var x = e.pageX - (boundingRect.left + document.body.scrollLeft);\n",
       "    var y = e.pageY - (boundingRect.top + document.body.scrollTop);\n",
       "\n",
       "    return { x: x, y: y };\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys(original) {\n",
       "    return Object.keys(original).reduce(function (obj, key) {\n",
       "        if (typeof original[key] !== 'object') {\n",
       "            obj[key] = original[key];\n",
       "        }\n",
       "        return obj;\n",
       "    }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function (event, name) {\n",
       "    var canvas_pos = mpl.findpos(event);\n",
       "\n",
       "    if (name === 'button_press') {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * this.ratio;\n",
       "    var y = canvas_pos.y * this.ratio;\n",
       "\n",
       "    this.send_message(name, {\n",
       "        x: x,\n",
       "        y: y,\n",
       "        button: event.button,\n",
       "        step: event.step,\n",
       "        guiEvent: simpleKeys(event),\n",
       "    });\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (_event, _name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.key_event = function (event, name) {\n",
       "    // Prevent repeat events\n",
       "    if (name === 'key_press') {\n",
       "        if (event.which === this._key) {\n",
       "            return;\n",
       "        } else {\n",
       "            this._key = event.which;\n",
       "        }\n",
       "    }\n",
       "    if (name === 'key_release') {\n",
       "        this._key = null;\n",
       "    }\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which !== 17) {\n",
       "        value += 'ctrl+';\n",
       "    }\n",
       "    if (event.altKey && event.which !== 18) {\n",
       "        value += 'alt+';\n",
       "    }\n",
       "    if (event.shiftKey && event.which !== 16) {\n",
       "        value += 'shift+';\n",
       "    }\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, { key: value, guiEvent: simpleKeys(event) });\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function (name) {\n",
       "    if (name === 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message('toolbar_button', { name: name });\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function (tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "\n",
       "///////////////// REMAINING CONTENT GENERATED BY embed_js.py /////////////////\n",
       "// prettier-ignore\n",
       "var _JSXTOOLS_RESIZE_OBSERVER=function(A){var t,i=new WeakMap,n=new WeakMap,a=new WeakMap,r=new WeakMap,o=new Set;function s(e){if(!(this instanceof s))throw new TypeError(\"Constructor requires 'new' operator\");i.set(this,e)}function h(){throw new TypeError(\"Function is not a constructor\")}function c(e,t,i,n){e=0 in arguments?Number(arguments[0]):0,t=1 in arguments?Number(arguments[1]):0,i=2 in arguments?Number(arguments[2]):0,n=3 in arguments?Number(arguments[3]):0,this.right=(this.x=this.left=e)+(this.width=i),this.bottom=(this.y=this.top=t)+(this.height=n),Object.freeze(this)}function d(){t=requestAnimationFrame(d);var s=new WeakMap,p=new Set;o.forEach((function(t){r.get(t).forEach((function(i){var r=t instanceof window.SVGElement,o=a.get(t),d=r?0:parseFloat(o.paddingTop),f=r?0:parseFloat(o.paddingRight),l=r?0:parseFloat(o.paddingBottom),u=r?0:parseFloat(o.paddingLeft),g=r?0:parseFloat(o.borderTopWidth),m=r?0:parseFloat(o.borderRightWidth),w=r?0:parseFloat(o.borderBottomWidth),b=u+f,F=d+l,v=(r?0:parseFloat(o.borderLeftWidth))+m,W=g+w,y=r?0:t.offsetHeight-W-t.clientHeight,E=r?0:t.offsetWidth-v-t.clientWidth,R=b+v,z=F+W,M=r?t.width:parseFloat(o.width)-R-E,O=r?t.height:parseFloat(o.height)-z-y;if(n.has(t)){var k=n.get(t);if(k[0]===M&&k[1]===O)return}n.set(t,[M,O]);var S=Object.create(h.prototype);S.target=t,S.contentRect=new c(u,d,M,O),s.has(i)||(s.set(i,[]),p.add(i)),s.get(i).push(S)}))})),p.forEach((function(e){i.get(e).call(e,s.get(e),e)}))}return s.prototype.observe=function(i){if(i instanceof window.Element){r.has(i)||(r.set(i,new Set),o.add(i),a.set(i,window.getComputedStyle(i)));var n=r.get(i);n.has(this)||n.add(this),cancelAnimationFrame(t),t=requestAnimationFrame(d)}},s.prototype.unobserve=function(i){if(i instanceof window.Element&&r.has(i)){var n=r.get(i);n.has(this)&&(n.delete(this),n.size||(r.delete(i),o.delete(i))),n.size||r.delete(i),o.size||cancelAnimationFrame(t)}},A.DOMRectReadOnly=c,A.ResizeObserver=s,A.ResizeObserverEntry=h,A}; // eslint-disable-line\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Left button pans, Right button zooms\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";/* global mpl */\n",
       "\n",
       "var comm_websocket_adapter = function (comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function () {\n",
       "        comm.close();\n",
       "    };\n",
       "    ws.send = function (m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function (msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data']);\n",
       "    });\n",
       "    return ws;\n",
       "};\n",
       "\n",
       "mpl.mpl_figure_comm = function (comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = document.getElementById(id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm);\n",
       "\n",
       "    function ondownload(figure, _format) {\n",
       "        window.open(figure.canvas.toDataURL());\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy, ondownload, element);\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element;\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error('Failed to find cell for figure', id, fig);\n",
       "        return;\n",
       "    }\n",
       "    fig.cell_info[0].output_area.element.on(\n",
       "        'cleared',\n",
       "        { fig: fig },\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function (fig, msg) {\n",
       "    var width = fig.canvas.width / fig.ratio;\n",
       "    fig.cell_info[0].output_area.element.off(\n",
       "        'cleared',\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "    fig.resizeObserverInstance.unobserve(fig.canvas_div);\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable();\n",
       "    fig.parent_element.innerHTML =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "    fig.close_ws(fig, msg);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.close_ws = function (fig, msg) {\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function (_remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width / this.ratio;\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message('ack', {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () {\n",
       "        fig.push_to_output();\n",
       "    }, 1000);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'btn-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'btn-group';\n",
       "    var button;\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'btn-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        button = fig.buttons[name] = document.createElement('button');\n",
       "        button.classList = 'btn btn-default';\n",
       "        button.href = '#';\n",
       "        button.title = name;\n",
       "        button.innerHTML = '<i class=\"fa ' + image + ' fa-lg\"></i>';\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message pull-right';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = document.createElement('div');\n",
       "    buttongrp.classList = 'btn-group inline pull-right';\n",
       "    button = document.createElement('button');\n",
       "    button.classList = 'btn btn-mini btn-primary';\n",
       "    button.href = '#';\n",
       "    button.title = 'Stop Interaction';\n",
       "    button.innerHTML = '<i class=\"fa fa-power-off icon-remove icon-large\"></i>';\n",
       "    button.addEventListener('click', function (_evt) {\n",
       "        fig.handle_close(fig, {});\n",
       "    });\n",
       "    button.addEventListener(\n",
       "        'mouseover',\n",
       "        on_mouseover_closure('Stop Interaction')\n",
       "    );\n",
       "    buttongrp.appendChild(button);\n",
       "    var titlebar = this.root.querySelector('.ui-dialog-titlebar');\n",
       "    titlebar.insertBefore(buttongrp, titlebar.firstChild);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._remove_fig_handler = function (event) {\n",
       "    var fig = event.data.fig;\n",
       "    if (event.target !== this) {\n",
       "        // Ignore bubbled events from children.\n",
       "        return;\n",
       "    }\n",
       "    fig.close_ws(fig, {});\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (el) {\n",
       "    el.style.boxSizing = 'content-box'; // override notebook setting of border-box.\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (el) {\n",
       "    // this is important to make the div 'focusable\n",
       "    el.setAttribute('tabindex', 0);\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    } else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (event, _name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager) {\n",
       "        manager = IPython.keyboard_manager;\n",
       "    }\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which === 13) {\n",
       "        this.canvas_div.blur();\n",
       "        // select the cell after this one\n",
       "        var index = IPython.notebook.find_cell_index(this.cell_info[0]);\n",
       "        IPython.notebook.select(index + 1);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "};\n",
       "\n",
       "mpl.find_output_cell = function (html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i = 0; i < ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code') {\n",
       "            for (var j = 0; j < cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] === html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "};\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel !== null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target(\n",
       "        'matplotlib',\n",
       "        mpl.mpl_figure_comm\n",
       "    );\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"700\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0.98, 'rolling window correlation')"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r_window_size = 120\n",
    "# Interpolate missing data.\n",
    "df_interpolated = new_df.interpolate()\n",
    "# Compute rolling window synchrony\n",
    "rolling_r = new_df['Close'].rolling(window=r_window_size, center=True).corr(new_df['Close_ndaq'])\n",
    "f,ax=plt.subplots(2,1,figsize=(7,6),sharex=True)\n",
    "new_df.rolling(window=30,center=True).median().plot(ax=ax[0])\n",
    "ax[0].set(xlabel='Frame',ylabel='price')\n",
    "rolling_r.plot(ax=ax[1])\n",
    "ax[1].set(xlabel='Frame',ylabel='Pearson r')\n",
    "plt.suptitle(\"rolling window correlation\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "id": "medieval-disaster",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"av_unique = get_av_uniqueness_from_triple_barrier(triple_barrier_events, df['Close'],\\nnum_threads=3)\""
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "'''av_unique = get_av_uniqueness_from_triple_barrier(triple_barrier_events, df['Close'],\n",
    "num_threads=3)'''"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "id": "noble-military",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n"
     ]
    }
   ],
   "source": [
    "triple_barrier_events.dropna(inplace = True)\n",
    "\n",
    "\n",
    "\n",
    "print(type(triple_barrier_events))\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "id": "independent-friendship",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2021-12-23 15:21:09.310709 100.0% num_concurrent_events done after 0.1 minutes. Remaining 0.0 minutes.\n",
      "2021-12-23 15:21:16.077798 100.0% _apply_weight_by_return done after 0.11 minutes. Remaining 0.0 minutes.\n"
     ]
    }
   ],
   "source": [
    "## must have clean versions of triple_barrier_events and df['Close']\n",
    "## no nan's\n",
    "\n",
    "\n",
    "from mlfinlab.sample_weights import get_weights_by_return\n",
    "\n",
    "\n",
    "#barrier_events = pd.read_csv('FILE_PATH', index_col=0, parse_dates=[0,2])\n",
    "\n",
    "#close_prices = pd.read_csv('FILE_PATH', index_col=0, parse_dates=[0,2])\n",
    "\n",
    "sample_weights = get_weights_by_return(triple_barrier_events, df['Close'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "id": "molecular-repository",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2021-11-26 19:47:00    3.779767\n",
       "2021-11-26 20:14:00    0.836857\n",
       "2021-11-26 20:26:00    0.422201\n",
       "2021-11-26 20:31:00    1.353223\n",
       "2021-11-26 20:35:00    1.955440\n",
       "dtype: float64"
      ]
     },
     "execution_count": 164,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample_weights.tail(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "id": "light-vanilla",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2021-12-23 15:33:37.474448 100.0% num_concurrent_events done after 0.11 minutes. Remaining 0.0 minutes.\n",
      "2021-12-23 15:33:43.544699 100.0% _get_average_uniqueness done after 0.09 minutes. Remaining 0.0 minutes.\n"
     ]
    }
   ],
   "source": [
    "from mlfinlab.sample_weights import get_weights_by_time_decay\n",
    "\n",
    "\n",
    "#barrier_events = pd.read_csv('FILE_PATH', index_col=0, parse_dates=[0,2])\n",
    "\n",
    "#close_prices = pd.read_csv('FILE_PATH', index_col=0, parse_dates=[0,2])\n",
    "\n",
    "sample_weights = get_weights_by_time_decay(triple_barrier_events, df['Close'], decay = 0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "id": "registered-westminster",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2020-02-03 17:31:00    0.500075\n",
       "2020-02-03 17:33:00    0.500135\n",
       "2020-02-03 17:39:00    0.500224\n",
       "2020-02-03 17:42:00    0.500319\n",
       "2020-02-03 17:45:00    0.500403\n",
       "Name: tW, dtype: float64"
      ]
     },
     "execution_count": 166,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample_weights.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "id": "facial-manitoba",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2021-11-26 19:47:00    0.999671\n",
       "2021-11-26 20:14:00    0.999765\n",
       "2021-11-26 20:26:00    0.999825\n",
       "2021-11-26 20:31:00    0.999880\n",
       "2021-11-26 20:35:00    1.000000\n",
       "Name: tW, dtype: float64"
      ]
     },
     "execution_count": 167,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample_weights.tail(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "id": "nasty-capability",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.series.Series'>\n"
     ]
    }
   ],
   "source": [
    "print(type(df['Close']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "id": "fluid-measure",
   "metadata": {},
   "outputs": [],
   "source": [
    "from mlfinlab.features.fracdiff import frac_diff_ffd\n",
    "\n",
    "\n",
    "\n",
    "frac_diff_series = frac_diff_ffd(combined_df, 0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "id": "continuous-viking",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(173648, 10)"
      ]
     },
     "execution_count": 171,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "frac_diff_series.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "id": "accepted-swimming",
   "metadata": {},
   "outputs": [],
   "source": [
    "label_list = list(combined_df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "id": "liquid-surprise",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "/* global mpl */\n",
       "window.mpl = {};\n",
       "\n",
       "mpl.get_websocket_type = function () {\n",
       "    if (typeof WebSocket !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof MozWebSocket !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert(\n",
       "            'Your browser does not have WebSocket support. ' +\n",
       "                'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "                'Firefox 4 and 5 are also supported but you ' +\n",
       "                'have to enable WebSockets in about:config.'\n",
       "        );\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure = function (figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = this.ws.binaryType !== undefined;\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById('mpl-warnings');\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent =\n",
       "                'This browser does not support binary websocket messages. ' +\n",
       "                'Performance may be slow.';\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = document.createElement('div');\n",
       "    this.root.setAttribute('style', 'display: inline-block');\n",
       "    this._root_extra_style(this.root);\n",
       "\n",
       "    parent_element.appendChild(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen = function () {\n",
       "        fig.send_message('supports_binary', { value: fig.supports_binary });\n",
       "        fig.send_message('send_image_mode', {});\n",
       "        if (fig.ratio !== 1) {\n",
       "            fig.send_message('set_dpi_ratio', { dpi_ratio: fig.ratio });\n",
       "        }\n",
       "        fig.send_message('refresh', {});\n",
       "    };\n",
       "\n",
       "    this.imageObj.onload = function () {\n",
       "        if (fig.image_mode === 'full') {\n",
       "            // Full images could contain transparency (where diff images\n",
       "            // almost always do), so we need to clear the canvas so that\n",
       "            // there is no ghosting.\n",
       "            fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "        }\n",
       "        fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "    };\n",
       "\n",
       "    this.imageObj.onunload = function () {\n",
       "        fig.ws.close();\n",
       "    };\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_header = function () {\n",
       "    var titlebar = document.createElement('div');\n",
       "    titlebar.classList =\n",
       "        'ui-dialog-titlebar ui-widget-header ui-corner-all ui-helper-clearfix';\n",
       "    var titletext = document.createElement('div');\n",
       "    titletext.classList = 'ui-dialog-title';\n",
       "    titletext.setAttribute(\n",
       "        'style',\n",
       "        'width: 100%; text-align: center; padding: 3px;'\n",
       "    );\n",
       "    titlebar.appendChild(titletext);\n",
       "    this.root.appendChild(titlebar);\n",
       "    this.header = titletext;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (_canvas_div) {};\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = (this.canvas_div = document.createElement('div'));\n",
       "    canvas_div.setAttribute(\n",
       "        'style',\n",
       "        'border: 1px solid #ddd;' +\n",
       "            'box-sizing: content-box;' +\n",
       "            'clear: both;' +\n",
       "            'min-height: 1px;' +\n",
       "            'min-width: 1px;' +\n",
       "            'outline: 0;' +\n",
       "            'overflow: hidden;' +\n",
       "            'position: relative;' +\n",
       "            'resize: both;'\n",
       "    );\n",
       "\n",
       "    function on_keyboard_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.key_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    canvas_div.addEventListener(\n",
       "        'keydown',\n",
       "        on_keyboard_event_closure('key_press')\n",
       "    );\n",
       "    canvas_div.addEventListener(\n",
       "        'keyup',\n",
       "        on_keyboard_event_closure('key_release')\n",
       "    );\n",
       "\n",
       "    this._canvas_extra_style(canvas_div);\n",
       "    this.root.appendChild(canvas_div);\n",
       "\n",
       "    var canvas = (this.canvas = document.createElement('canvas'));\n",
       "    canvas.classList.add('mpl-canvas');\n",
       "    canvas.setAttribute('style', 'box-sizing: content-box;');\n",
       "\n",
       "    this.context = canvas.getContext('2d');\n",
       "\n",
       "    var backingStore =\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        this.context.webkitBackingStorePixelRatio ||\n",
       "        this.context.mozBackingStorePixelRatio ||\n",
       "        this.context.msBackingStorePixelRatio ||\n",
       "        this.context.oBackingStorePixelRatio ||\n",
       "        this.context.backingStorePixelRatio ||\n",
       "        1;\n",
       "\n",
       "    this.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband_canvas = (this.rubberband_canvas = document.createElement(\n",
       "        'canvas'\n",
       "    ));\n",
       "    rubberband_canvas.setAttribute(\n",
       "        'style',\n",
       "        'box-sizing: content-box; position: absolute; left: 0; top: 0; z-index: 1;'\n",
       "    );\n",
       "\n",
       "    // Apply a ponyfill if ResizeObserver is not implemented by browser.\n",
       "    if (this.ResizeObserver === undefined) {\n",
       "        if (window.ResizeObserver !== undefined) {\n",
       "            this.ResizeObserver = window.ResizeObserver;\n",
       "        } else {\n",
       "            var obs = _JSXTOOLS_RESIZE_OBSERVER({});\n",
       "            this.ResizeObserver = obs.ResizeObserver;\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.resizeObserverInstance = new this.ResizeObserver(function (entries) {\n",
       "        var nentries = entries.length;\n",
       "        for (var i = 0; i < nentries; i++) {\n",
       "            var entry = entries[i];\n",
       "            var width, height;\n",
       "            if (entry.contentBoxSize) {\n",
       "                if (entry.contentBoxSize instanceof Array) {\n",
       "                    // Chrome 84 implements new version of spec.\n",
       "                    width = entry.contentBoxSize[0].inlineSize;\n",
       "                    height = entry.contentBoxSize[0].blockSize;\n",
       "                } else {\n",
       "                    // Firefox implements old version of spec.\n",
       "                    width = entry.contentBoxSize.inlineSize;\n",
       "                    height = entry.contentBoxSize.blockSize;\n",
       "                }\n",
       "            } else {\n",
       "                // Chrome <84 implements even older version of spec.\n",
       "                width = entry.contentRect.width;\n",
       "                height = entry.contentRect.height;\n",
       "            }\n",
       "\n",
       "            // Keep the size of the canvas and rubber band canvas in sync with\n",
       "            // the canvas container.\n",
       "            if (entry.devicePixelContentBoxSize) {\n",
       "                // Chrome 84 implements new version of spec.\n",
       "                canvas.setAttribute(\n",
       "                    'width',\n",
       "                    entry.devicePixelContentBoxSize[0].inlineSize\n",
       "                );\n",
       "                canvas.setAttribute(\n",
       "                    'height',\n",
       "                    entry.devicePixelContentBoxSize[0].blockSize\n",
       "                );\n",
       "            } else {\n",
       "                canvas.setAttribute('width', width * fig.ratio);\n",
       "                canvas.setAttribute('height', height * fig.ratio);\n",
       "            }\n",
       "            canvas.setAttribute(\n",
       "                'style',\n",
       "                'width: ' + width + 'px; height: ' + height + 'px;'\n",
       "            );\n",
       "\n",
       "            rubberband_canvas.setAttribute('width', width);\n",
       "            rubberband_canvas.setAttribute('height', height);\n",
       "\n",
       "            // And update the size in Python. We ignore the initial 0/0 size\n",
       "            // that occurs as the element is placed into the DOM, which should\n",
       "            // otherwise not happen due to the minimum size styling.\n",
       "            if (fig.ws.readyState == 1 && width != 0 && height != 0) {\n",
       "                fig.request_resize(width, height);\n",
       "            }\n",
       "        }\n",
       "    });\n",
       "    this.resizeObserverInstance.observe(canvas_div);\n",
       "\n",
       "    function on_mouse_event_closure(name) {\n",
       "        return function (event) {\n",
       "            return fig.mouse_event(event, name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousedown',\n",
       "        on_mouse_event_closure('button_press')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseup',\n",
       "        on_mouse_event_closure('button_release')\n",
       "    );\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mousemove',\n",
       "        on_mouse_event_closure('motion_notify')\n",
       "    );\n",
       "\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseenter',\n",
       "        on_mouse_event_closure('figure_enter')\n",
       "    );\n",
       "    rubberband_canvas.addEventListener(\n",
       "        'mouseleave',\n",
       "        on_mouse_event_closure('figure_leave')\n",
       "    );\n",
       "\n",
       "    canvas_div.addEventListener('wheel', function (event) {\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        on_mouse_event_closure('scroll')(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.appendChild(canvas);\n",
       "    canvas_div.appendChild(rubberband_canvas);\n",
       "\n",
       "    this.rubberband_context = rubberband_canvas.getContext('2d');\n",
       "    this.rubberband_context.strokeStyle = '#000000';\n",
       "\n",
       "    this._resize_canvas = function (width, height, forward) {\n",
       "        if (forward) {\n",
       "            canvas_div.style.width = width + 'px';\n",
       "            canvas_div.style.height = height + 'px';\n",
       "        }\n",
       "    };\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    this.rubberband_canvas.addEventListener('contextmenu', function (_e) {\n",
       "        event.preventDefault();\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus() {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'mpl-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'mpl-button-group';\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'mpl-button-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        var button = (fig.buttons[name] = document.createElement('button'));\n",
       "        button.classList = 'mpl-widget';\n",
       "        button.setAttribute('role', 'button');\n",
       "        button.setAttribute('aria-disabled', 'false');\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "\n",
       "        var icon_img = document.createElement('img');\n",
       "        icon_img.src = '_images/' + image + '.png';\n",
       "        icon_img.srcset = '_images/' + image + '_large.png 2x';\n",
       "        icon_img.alt = tooltip;\n",
       "        button.appendChild(icon_img);\n",
       "\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    var fmt_picker = document.createElement('select');\n",
       "    fmt_picker.classList = 'mpl-widget';\n",
       "    toolbar.appendChild(fmt_picker);\n",
       "    this.format_dropdown = fmt_picker;\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = document.createElement('option');\n",
       "        option.selected = fmt === mpl.default_extension;\n",
       "        option.innerHTML = fmt;\n",
       "        fmt_picker.appendChild(option);\n",
       "    }\n",
       "\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.request_resize = function (x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', { width: x_pixels, height: y_pixels });\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_message = function (type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function () {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({ type: 'draw', figure_id: this.id }));\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function (fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] !== fig.canvas.width || size[1] !== fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1], msg['forward']);\n",
       "        fig.send_message('refresh', {});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function (fig, msg) {\n",
       "    var x0 = msg['x0'] / fig.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / fig.ratio;\n",
       "    var x1 = msg['x1'] / fig.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / fig.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0,\n",
       "        0,\n",
       "        fig.canvas.width / fig.ratio,\n",
       "        fig.canvas.height / fig.ratio\n",
       "    );\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function (fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function (fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch (cursor) {\n",
       "        case 0:\n",
       "            cursor = 'pointer';\n",
       "            break;\n",
       "        case 1:\n",
       "            cursor = 'default';\n",
       "            break;\n",
       "        case 2:\n",
       "            cursor = 'crosshair';\n",
       "            break;\n",
       "        case 3:\n",
       "            cursor = 'move';\n",
       "            break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_message = function (fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function (fig, _msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function (fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_history_buttons = function (fig, msg) {\n",
       "    for (var key in msg) {\n",
       "        if (!(key in fig.buttons)) {\n",
       "            continue;\n",
       "        }\n",
       "        fig.buttons[key].disabled = !msg[key];\n",
       "        fig.buttons[key].setAttribute('aria-disabled', !msg[key]);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_navigate_mode = function (fig, msg) {\n",
       "    if (msg['mode'] === 'PAN') {\n",
       "        fig.buttons['Pan'].classList.add('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    } else if (msg['mode'] === 'ZOOM') {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.add('active');\n",
       "    } else {\n",
       "        fig.buttons['Pan'].classList.remove('active');\n",
       "        fig.buttons['Zoom'].classList.remove('active');\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message('ack', {});\n",
       "};\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function (fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = 'image/png';\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src\n",
       "                );\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data\n",
       "            );\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        } else if (\n",
       "            typeof evt.data === 'string' &&\n",
       "            evt.data.slice(0, 21) === 'data:image/png;base64'\n",
       "        ) {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig['handle_' + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\n",
       "                \"No handler for the '\" + msg_type + \"' message type: \",\n",
       "                msg\n",
       "            );\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\n",
       "                    \"Exception inside the 'handler_\" + msg_type + \"' callback:\",\n",
       "                    e,\n",
       "                    e.stack,\n",
       "                    msg\n",
       "                );\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "};\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function (e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e) {\n",
       "        e = window.event;\n",
       "    }\n",
       "    if (e.target) {\n",
       "        targ = e.target;\n",
       "    } else if (e.srcElement) {\n",
       "        targ = e.srcElement;\n",
       "    }\n",
       "    if (targ.nodeType === 3) {\n",
       "        // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "    }\n",
       "\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    var boundingRect = targ.getBoundingClientRect();\n",
       "    var x = e.pageX - (boundingRect.left + document.body.scrollLeft);\n",
       "    var y = e.pageY - (boundingRect.top + document.body.scrollTop);\n",
       "\n",
       "    return { x: x, y: y };\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys(original) {\n",
       "    return Object.keys(original).reduce(function (obj, key) {\n",
       "        if (typeof original[key] !== 'object') {\n",
       "            obj[key] = original[key];\n",
       "        }\n",
       "        return obj;\n",
       "    }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function (event, name) {\n",
       "    var canvas_pos = mpl.findpos(event);\n",
       "\n",
       "    if (name === 'button_press') {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * this.ratio;\n",
       "    var y = canvas_pos.y * this.ratio;\n",
       "\n",
       "    this.send_message(name, {\n",
       "        x: x,\n",
       "        y: y,\n",
       "        button: event.button,\n",
       "        step: event.step,\n",
       "        guiEvent: simpleKeys(event),\n",
       "    });\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (_event, _name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.key_event = function (event, name) {\n",
       "    // Prevent repeat events\n",
       "    if (name === 'key_press') {\n",
       "        if (event.which === this._key) {\n",
       "            return;\n",
       "        } else {\n",
       "            this._key = event.which;\n",
       "        }\n",
       "    }\n",
       "    if (name === 'key_release') {\n",
       "        this._key = null;\n",
       "    }\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which !== 17) {\n",
       "        value += 'ctrl+';\n",
       "    }\n",
       "    if (event.altKey && event.which !== 18) {\n",
       "        value += 'alt+';\n",
       "    }\n",
       "    if (event.shiftKey && event.which !== 16) {\n",
       "        value += 'shift+';\n",
       "    }\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, { key: value, guiEvent: simpleKeys(event) });\n",
       "    return false;\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function (name) {\n",
       "    if (name === 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message('toolbar_button', { name: name });\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function (tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "\n",
       "///////////////// REMAINING CONTENT GENERATED BY embed_js.py /////////////////\n",
       "// prettier-ignore\n",
       "var _JSXTOOLS_RESIZE_OBSERVER=function(A){var t,i=new WeakMap,n=new WeakMap,a=new WeakMap,r=new WeakMap,o=new Set;function s(e){if(!(this instanceof s))throw new TypeError(\"Constructor requires 'new' operator\");i.set(this,e)}function h(){throw new TypeError(\"Function is not a constructor\")}function c(e,t,i,n){e=0 in arguments?Number(arguments[0]):0,t=1 in arguments?Number(arguments[1]):0,i=2 in arguments?Number(arguments[2]):0,n=3 in arguments?Number(arguments[3]):0,this.right=(this.x=this.left=e)+(this.width=i),this.bottom=(this.y=this.top=t)+(this.height=n),Object.freeze(this)}function d(){t=requestAnimationFrame(d);var s=new WeakMap,p=new Set;o.forEach((function(t){r.get(t).forEach((function(i){var r=t instanceof window.SVGElement,o=a.get(t),d=r?0:parseFloat(o.paddingTop),f=r?0:parseFloat(o.paddingRight),l=r?0:parseFloat(o.paddingBottom),u=r?0:parseFloat(o.paddingLeft),g=r?0:parseFloat(o.borderTopWidth),m=r?0:parseFloat(o.borderRightWidth),w=r?0:parseFloat(o.borderBottomWidth),b=u+f,F=d+l,v=(r?0:parseFloat(o.borderLeftWidth))+m,W=g+w,y=r?0:t.offsetHeight-W-t.clientHeight,E=r?0:t.offsetWidth-v-t.clientWidth,R=b+v,z=F+W,M=r?t.width:parseFloat(o.width)-R-E,O=r?t.height:parseFloat(o.height)-z-y;if(n.has(t)){var k=n.get(t);if(k[0]===M&&k[1]===O)return}n.set(t,[M,O]);var S=Object.create(h.prototype);S.target=t,S.contentRect=new c(u,d,M,O),s.has(i)||(s.set(i,[]),p.add(i)),s.get(i).push(S)}))})),p.forEach((function(e){i.get(e).call(e,s.get(e),e)}))}return s.prototype.observe=function(i){if(i instanceof window.Element){r.has(i)||(r.set(i,new Set),o.add(i),a.set(i,window.getComputedStyle(i)));var n=r.get(i);n.has(this)||n.add(this),cancelAnimationFrame(t),t=requestAnimationFrame(d)}},s.prototype.unobserve=function(i){if(i instanceof window.Element&&r.has(i)){var n=r.get(i);n.has(this)&&(n.delete(this),n.size||(r.delete(i),o.delete(i))),n.size||r.delete(i),o.size||cancelAnimationFrame(t)}},A.DOMRectReadOnly=c,A.ResizeObserver=s,A.ResizeObserverEntry=h,A}; // eslint-disable-line\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Left button pans, Right button zooms\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\\nx/y fixes axis, CTRL fixes aspect\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";/* global mpl */\n",
       "\n",
       "var comm_websocket_adapter = function (comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function () {\n",
       "        comm.close();\n",
       "    };\n",
       "    ws.send = function (m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function (msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data']);\n",
       "    });\n",
       "    return ws;\n",
       "};\n",
       "\n",
       "mpl.mpl_figure_comm = function (comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = document.getElementById(id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm);\n",
       "\n",
       "    function ondownload(figure, _format) {\n",
       "        window.open(figure.canvas.toDataURL());\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy, ondownload, element);\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element;\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error('Failed to find cell for figure', id, fig);\n",
       "        return;\n",
       "    }\n",
       "    fig.cell_info[0].output_area.element.on(\n",
       "        'cleared',\n",
       "        { fig: fig },\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function (fig, msg) {\n",
       "    var width = fig.canvas.width / fig.ratio;\n",
       "    fig.cell_info[0].output_area.element.off(\n",
       "        'cleared',\n",
       "        fig._remove_fig_handler\n",
       "    );\n",
       "    fig.resizeObserverInstance.unobserve(fig.canvas_div);\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable();\n",
       "    fig.parent_element.innerHTML =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "    fig.close_ws(fig, msg);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.close_ws = function (fig, msg) {\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function (_remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width / this.ratio;\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] =\n",
       "        '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function () {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message('ack', {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () {\n",
       "        fig.push_to_output();\n",
       "    }, 1000);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function () {\n",
       "    var fig = this;\n",
       "\n",
       "    var toolbar = document.createElement('div');\n",
       "    toolbar.classList = 'btn-toolbar';\n",
       "    this.root.appendChild(toolbar);\n",
       "\n",
       "    function on_click_closure(name) {\n",
       "        return function (_event) {\n",
       "            return fig.toolbar_button_onclick(name);\n",
       "        };\n",
       "    }\n",
       "\n",
       "    function on_mouseover_closure(tooltip) {\n",
       "        return function (event) {\n",
       "            if (!event.currentTarget.disabled) {\n",
       "                return fig.toolbar_button_onmouseover(tooltip);\n",
       "            }\n",
       "        };\n",
       "    }\n",
       "\n",
       "    fig.buttons = {};\n",
       "    var buttonGroup = document.createElement('div');\n",
       "    buttonGroup.classList = 'btn-group';\n",
       "    var button;\n",
       "    for (var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            /* Instead of a spacer, we start a new button group. */\n",
       "            if (buttonGroup.hasChildNodes()) {\n",
       "                toolbar.appendChild(buttonGroup);\n",
       "            }\n",
       "            buttonGroup = document.createElement('div');\n",
       "            buttonGroup.classList = 'btn-group';\n",
       "            continue;\n",
       "        }\n",
       "\n",
       "        button = fig.buttons[name] = document.createElement('button');\n",
       "        button.classList = 'btn btn-default';\n",
       "        button.href = '#';\n",
       "        button.title = name;\n",
       "        button.innerHTML = '<i class=\"fa ' + image + ' fa-lg\"></i>';\n",
       "        button.addEventListener('click', on_click_closure(method_name));\n",
       "        button.addEventListener('mouseover', on_mouseover_closure(tooltip));\n",
       "        buttonGroup.appendChild(button);\n",
       "    }\n",
       "\n",
       "    if (buttonGroup.hasChildNodes()) {\n",
       "        toolbar.appendChild(buttonGroup);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = document.createElement('span');\n",
       "    status_bar.classList = 'mpl-message pull-right';\n",
       "    toolbar.appendChild(status_bar);\n",
       "    this.message = status_bar;\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = document.createElement('div');\n",
       "    buttongrp.classList = 'btn-group inline pull-right';\n",
       "    button = document.createElement('button');\n",
       "    button.classList = 'btn btn-mini btn-primary';\n",
       "    button.href = '#';\n",
       "    button.title = 'Stop Interaction';\n",
       "    button.innerHTML = '<i class=\"fa fa-power-off icon-remove icon-large\"></i>';\n",
       "    button.addEventListener('click', function (_evt) {\n",
       "        fig.handle_close(fig, {});\n",
       "    });\n",
       "    button.addEventListener(\n",
       "        'mouseover',\n",
       "        on_mouseover_closure('Stop Interaction')\n",
       "    );\n",
       "    buttongrp.appendChild(button);\n",
       "    var titlebar = this.root.querySelector('.ui-dialog-titlebar');\n",
       "    titlebar.insertBefore(buttongrp, titlebar.firstChild);\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._remove_fig_handler = function (event) {\n",
       "    var fig = event.data.fig;\n",
       "    if (event.target !== this) {\n",
       "        // Ignore bubbled events from children.\n",
       "        return;\n",
       "    }\n",
       "    fig.close_ws(fig, {});\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function (el) {\n",
       "    el.style.boxSizing = 'content-box'; // override notebook setting of border-box.\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function (el) {\n",
       "    // this is important to make the div 'focusable\n",
       "    el.setAttribute('tabindex', 0);\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    } else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function (event, _name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager) {\n",
       "        manager = IPython.keyboard_manager;\n",
       "    }\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which === 13) {\n",
       "        this.canvas_div.blur();\n",
       "        // select the cell after this one\n",
       "        var index = IPython.notebook.find_cell_index(this.cell_info[0]);\n",
       "        IPython.notebook.select(index + 1);\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_save = function (fig, _msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "};\n",
       "\n",
       "mpl.find_output_cell = function (html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i = 0; i < ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code') {\n",
       "            for (var j = 0; j < cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] === html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "};\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel !== null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target(\n",
       "        'matplotlib',\n",
       "        mpl.mpl_figure_comm\n",
       "    );\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"640\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x18093df3790>"
      ]
     },
     "execution_count": 178,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%matplotlib notebook\n",
    "\n",
    "plt.plot(frac_diff_series)\n",
    "plt.legend(label_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "amended-spirit",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "seasonal-joyce",
   "metadata": {},
   "outputs": [],
   "source": [
    "import full_export_tools_original as tools\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "accessible-hometown",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "I work good\n"
     ]
    }
   ],
   "source": [
    "tools.test_import()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "architectural-parking",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "stable-frank",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "exotic-treatment",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "devoted-masters",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "organized-carter",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "conscious-forwarding",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "valid-program",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bronze-witness",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "palestinian-elements",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "adult-spring",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "suspended-threshold",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "referenced-grill",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "danish-boards",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "chinese-profile",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "found-withdrawal",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "duplicate-windows",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "grand-crowd",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "creative-declaration",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "handed-machinery",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "received-therapist",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bulgarian-mambo",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "functioning-panic",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "coated-thomas",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "mineral-resident",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bronze-lodging",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "corporate-equity",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "raised-warrant",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "mechanical-survivor",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "romantic-greece",
   "metadata": {},
   "outputs": [],
   "source": [
    "from fracdiff import FracdiffStat\n",
    "from fracdiff import Fracdiff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "suspected-orleans",
   "metadata": {},
   "outputs": [],
   "source": [
    "from autoviz.AutoViz_Class import AutoViz_Class\n",
    "AV = AutoViz_Class()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "twelve-dominican",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(df.columns.to_list())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "human-plaza",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "placed-encounter",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_frac_diff = df.iloc[:,:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "precious-defendant",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_frac_diff = test_frac_diff.values.reshape(-1,test_frac_diff.shape[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "seven-western",
   "metadata": {},
   "outputs": [],
   "source": [
    "f = FracdiffStat(window = 30)\n",
    "frac_diff_out = f.fit_transform(test_frac_diff)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "technological-edmonton",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "touched-street",
   "metadata": {},
   "outputs": [],
   "source": [
    "#f = Fracdiff(d = d_val, window = 60)\n",
    "#frac_diff_out = f.transform(test_frac_diff)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "metallic-rebecca",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_frac = pd.DataFrame(frac_diff_out, index = df.index, columns = df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "accepted-romantic",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "authorized-amino",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_frac.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "arranged-pocket",
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib notebook\n",
    "plt.plot(df_frac['Close'])\n",
    "plt.plot(adj_close_list)\n",
    "plt.legend(['Close frac','true'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "contemporary-speaker",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.rename({\"Close\": \"Adj Close\"}, axis = 1, inplace = True)\n",
    "\n",
    "df.head(4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ignored-lesson",
   "metadata": {},
   "outputs": [],
   "source": [
    "'''#2 min ticker\n",
    "# 30 intervals = 1 hour << OLD\n",
    "# 195 intervals = trading day'''\n",
    "                                            # there are more intervals that we can use / change\n",
    "    \n",
    "# 390 intervals / day < 1 min intervals\n",
    "# 390 < 1 day, 780 < 2 day, 1170 < 3 days, 1560 < 4 day, 1950 < 5 days\n",
    "# 2340 < 6 days, 2730 < 7 days, 3120 < 8 days, 3510 < 9 days,\n",
    "# 3900 < 10 days, 4290 < 11 days, 4680 < 12 days, 5070 < 13 days,\n",
    "# 5460 < 14 days, 5850 < 15 days, 6240 < 16 days, 6630 < 17 days,\n",
    "\n",
    "#min tickers\n",
    "#\n",
    "really_fast = 5\n",
    "fast = 30\n",
    "slow = 120\n",
    "\n",
    "add_indicators()\n",
    "\n",
    "# Convert Date column to datetime\n",
    "df.reset_index(level=0, inplace=True)\n",
    "\n",
    "df.columns = [str(x).lower().replace(' ', '_') for x in df.columns]\n",
    "\n",
    "# Change all column headings to be lower case, and remove spacing\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "simplified-ambassador",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.columns.to_list()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bulgarian-belly",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.tail(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "greenhouse-storm",
   "metadata": {},
   "outputs": [],
   "source": [
    "#print(df.columns.to_list())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "republican-starter",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_frac.rename({\"Close\": \"Adj Close\"}, axis = 1, inplace = True)\n",
    "\n",
    "df_frac.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "greatest-assumption",
   "metadata": {},
   "outputs": [],
   "source": [
    "df[['open', 'high', 'low','adj_close']] = df_frac[['Open', 'High', 'Low', 'Adj Close']].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "floating-riding",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.tail(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "beginning-sherman",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get difference between high and low of each day\n",
    "df['range_hl'] = df['high'] - df['low']\n",
    "df.drop(['high', 'low'], axis=1, inplace=True)\n",
    "# Get difference between open and close of each day\n",
    "df['range_oc'] = df['open'] - df['adj_close']\n",
    "df.drop(['open'], axis=1, inplace=True)\n",
    "# Add a column 'order_day' to indicate the order of the rows by date\n",
    "df['order_day'] = [x for x in list(range(len(df)))]\n",
    "# merging_keys\n",
    "merging_keys = ['order_day']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "inclusive-designation",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "legitimate-orleans",
   "metadata": {},
   "outputs": [],
   "source": [
    "num_interval_lag = 30\n",
    "\n",
    "df = add_lag(num_interval_lag, df)\n",
    "\n",
    "#df['adj_close'] = df['adj_close'].shift(30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "upper-saint",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.fillna(0, inplace=True)\n",
    "\n",
    "#df.tail(35)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "aquatic-residence",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.columns.to_list()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "differential-desert",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.info(memory_usage = 'deep')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "earlier-stadium",
   "metadata": {},
   "outputs": [],
   "source": [
    "types_list = df.dtypes\n",
    "types_list.to_dict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "economic-honduras",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "df = df.select_dtypes(include = ['float64']).astype('float16', copy = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "alternate-campaign",
   "metadata": {},
   "outputs": [],
   "source": [
    "types_list = df.dtypes\n",
    "types_list.to_dict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "thorough-coaching",
   "metadata": {},
   "outputs": [],
   "source": [
    "df['datetime'] = df_frac.index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "published-voltage",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.info(memory_usage = 'deep')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "numerous-emission",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.tail(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "broadband-serial",
   "metadata": {},
   "outputs": [],
   "source": [
    "'''7020\t18\n",
    "7410\t19\n",
    "7800\t20\n",
    "8190\t21\n",
    "8580\t22\n",
    "8970\t23\n",
    "9360\t24\n",
    "9750\t25\n",
    "10140\t26\n",
    "10530\t27\n",
    "10920\t28\n",
    "11310\t29\n",
    "11700\t30\n",
    "12090\t31\n",
    "12480\t32\n",
    "12870\t33\n",
    "13260\t34\n",
    "13650\t35\n",
    "14040\t36\n",
    "14430\t37\n",
    "14820\t38\n",
    "15210\t39\n",
    "15600\t40\n",
    "15990\t41\n",
    "16380\t42\n",
    "16770\t43\n",
    "17160\t44\n",
    "17550\t45\n",
    "17940\t46\n",
    "18330\t47\n",
    "18720\t48\n",
    "19110\t49\n",
    "19500\t50\n",
    "19890\t51\n",
    "20280\t52\n",
    "20670\t53\n",
    "21060\t54\n",
    "21450\t55\n",
    "21840\t56\n",
    "22230\t57\n",
    "22620\t58\n",
    "23010\t59\n",
    "23400\t60\n",
    "23790\t61\n",
    "24180\t62\n",
    "24570\t63\n",
    "24960\t64\n",
    "25350\t65\n",
    "25740\t66\n",
    "26130\t67\n",
    "26520\t68\n",
    "26910\t69\n",
    "27300\t70\n",
    "27690\t71\n",
    "28080\t72\n",
    "28470\t73\n",
    "28860\t74\n",
    "29250\t75\n",
    "29640\t76\n",
    "30030\t77\n",
    "30420\t78\n",
    "30810\t79\n",
    "31200\t80\n",
    "31590\t81\n",
    "31980\t82\n",
    "32370\t83\n",
    "32760\t84\n",
    "33150\t85\n",
    "33540\t86\n",
    "33930\t87\n",
    "34320\t88\n",
    "34710\t89\n",
    "35100\t90'''\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "processed-hollywood",
   "metadata": {},
   "outputs": [],
   "source": [
    "#add_labels(df)\n",
    "\n",
    "\n",
    "# 390 intervals / day < 1 min intervals\n",
    "# 390 < 1 day, 780 < 2 day, 1170 < 3 days, 1560 < 4 day, 1950 < 5 days\n",
    "# 2340 < 6 days, 2730 < 7 days, 3120 < 8 days, 3510 < 9 days,\n",
    "# 3900 < 10 days, 4290 < 11 days, 4680 < 12 days, 5070 < 13 days,\n",
    "# 5460 < 14 days, 5850 < 15 days, 6240 < 16 days, 6630 < 17 days,\n",
    "\n",
    "window_size = 15\n",
    "add_labels(df, window_size)\n",
    "\n",
    "\n",
    "#print(df['labels'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "renewable-catalyst",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.tail(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "illegal-robinson",
   "metadata": {},
   "outputs": [],
   "source": [
    "buy_true_markers = df.index[df['labels']==1].tolist()\n",
    "sell_true_markers = df.index[df['labels']==0].tolist()\n",
    "#BUY => 1, SELL => 0, HOLD => 2\n",
    "all_markers = df['labels']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "discrete-brazil",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "direct-region",
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib notebook\n",
    "plt.rcParams['figure.figsize'] = [10, 6]\n",
    "plt.plot(df['datetime'], adj_close_list)\n",
    "plt.plot(df['datetime'], adj_close_list, 'g^', markevery = buy_true_markers)\n",
    "plt.plot(df['datetime'], adj_close_list, 'rv', markevery = sell_true_markers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "chronic-indonesian",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "pending-campbell",
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib notebook\n",
    "\n",
    "plt.plot(df['datetime'], adj_close_list)\n",
    "plt.plot(df['datetime'], df['upper_band_nine'], 'r:')\n",
    "plt.plot(df['datetime'], df['lower_band_nine'], 'g:')\n",
    "plt.plot(df['datetime'], adj_close_list, 'g^', markevery = buy_true_markers)\n",
    "plt.plot(df['datetime'], adj_close_list, 'rv', markevery = sell_true_markers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "hearing-ceramic",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "capable-cuisine",
   "metadata": {},
   "outputs": [],
   "source": [
    "label_vals = df['labels'].value_counts()\n",
    "df['labels'].value_counts()\n",
    "#BUY => 1, SELL => 0, HOLD => 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "automotive-surgery",
   "metadata": {},
   "outputs": [],
   "source": [
    "num_buys = label_vals[1]\n",
    "num_sells = label_vals[0]\n",
    "num_holds = label_vals[2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "expressed-reggae",
   "metadata": {},
   "outputs": [],
   "source": [
    "least_vals = min(num_buys, num_sells, num_holds)\n",
    "#print(least_vals)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "important-greenhouse",
   "metadata": {},
   "outputs": [],
   "source": [
    "imp_factor = 5\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "willing-soviet",
   "metadata": {},
   "outputs": [],
   "source": [
    "weight_buys = imp_factor * (least_vals / num_buys)\n",
    "weight_sells = imp_factor * (least_vals / num_sells)\n",
    "weight_holds = least_vals / num_holds\n",
    "print(weight_buys)\n",
    "print(weight_sells)\n",
    "print(weight_holds)\n",
    "#BUY => 1, SELL => 0, HOLD => 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "combined-prophet",
   "metadata": {},
   "outputs": [],
   "source": [
    "buy_list = df.index[df['labels'] == 1].tolist()\n",
    "sell_list = df.index[df['labels'] == 0].tolist()\n",
    "hold_list = df.index[df['labels'] == 2].tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "theoretical-shaft",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.loc[buy_list, 'weights'] = weight_buys\n",
    "df.loc[sell_list, 'weights'] = weight_sells\n",
    "df.loc[hold_list, 'weights'] = weight_holds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "unsigned-profile",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "electronic-blowing",
   "metadata": {},
   "outputs": [],
   "source": [
    "#df['weights'].tail(50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "joined-influence",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "periodic-staff",
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_to_rem = ['weights','adj_close','ema', 'slowk','slowd','r_s_i','fastk','fastd','williamsr','volume','range_hl','range_oc', 'obv', 'adosc', 'ogchaikin', 'htdcperiod','htdcphase',\n",
    "                'inphase','quad','rsin','leadsin', 'two_crows', 'three_crows', 'three_inside', 'three_line', 'three_out', 'three_stars', 'three_soldier', 'baby', 'adv', 'belt_hold',\n",
    "                'breakaway', 'closingmara', 'baby_swallow', 'counter','dark_cloud','doji','doji_star','dragon_doji','engulf','evening_star','gapside','gravestone','hammer',\n",
    "                'hang_man','harami','harami_cross','high_wave','hikkake','hikkake_mod','pidgeon','id_three_crows','in_neck','inv_hammer','kicking','kicking_len','ladder_bot',\n",
    "                'doji_long','long_line','marabozu', 'match_glow','mat_hold','morning_doji','morning_star','on_neck','pierce','rickshaw','rise_fall','sep_line','shooting_star',\n",
    "                'sl_candle','spin_top','stalled','stick_sand','takuri','tasuki_gap','thrust','tristar','three_river','ud_two_gap','down_three_gap', 'upper_band','lower_band',\n",
    "                'mid_band','d_ema','ht_trend','kama','ma_fast','ma_really_fast','ma_slow','mid','mid_price','sar','sarext','sma','tema','trima','wma','adx','adxr','apo','aroon_d','aroon_u','aroon_osc',\n",
    "                'bop','cci','cmo','dx','macd','macdsig','macdhist','macdex','macdexsig','macdexhist','macdfixd','macdfixdsig','macdfixdhist','mfi','min_di','min_dm',\n",
    "                'momo','plus_di','plus_dm','ppo','roc','rocp','rocr','rocr_hund','rsi_fastk','rsi_fastd','trix','ult_osc', 'atr','natr','t_range',\n",
    "                'upper_band_nine','lower_band_nine','mid_band_nine','upper_band_one','lower_band_one','mid_band_one','upper_band_slow','lower_band_slow','mid_band_slow' \n",
    "                ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "intended-ottawa",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "spiritual-comfort",
   "metadata": {},
   "outputs": [],
   "source": [
    "close_vals_list = df['adj_close']\n",
    "dates_list = df['datetime']\n",
    "weights_list = df['weights']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "velvet-chocolate",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fancy-width",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(df.columns.to_list())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "congressional-chicago",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "progressive-situation",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.drop(cols_to_rem, axis = 1, inplace = True, errors = 'ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "available-departure",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_cols = df.columns.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "australian-console",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "complicated-richards",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(df.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cross-overall",
   "metadata": {},
   "outputs": [],
   "source": [
    "#print(weights_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "governing-heath",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "#BUY => 1, SELL => 0, HOLD => 2\n",
    "\n",
    "num_test = int(.2*len(df))\n",
    "num_train = len(df) - num_test\n",
    "\n",
    "\n",
    "print(df['labels'].iloc[[-1]])\n",
    "\n",
    "#train_close_vals = close_vals_list[:num_train]\n",
    "#test_close_vals = close_vals_list[num_train:]\n",
    "\n",
    "#test_dates_list = test['date']\n",
    "#BUY => 1, SELL => 0, HOLD => 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "revolutionary-operator",
   "metadata": {},
   "outputs": [],
   "source": [
    "df['labels'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "african-limitation",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dimensional-search",
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_to_scale = add_scale(num_interval_lag)\n",
    "\n",
    "#cols_to_scale.remove(cols_to_rem)\n",
    "\n",
    "# Do scaling for train set\n",
    "# Here we only scale the train dataset, and not the entire dataset to prevent information leak\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "steady-present",
   "metadata": {},
   "outputs": [],
   "source": [
    "for element in cols_to_rem:\n",
    "\n",
    "    if element in cols_to_scale:\n",
    "\n",
    "        cols_to_scale.remove(element)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "sublime-shepherd",
   "metadata": {},
   "outputs": [],
   "source": [
    "len(cols_to_scale)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "after-climb",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df[slow:]\n",
    "weights_list = weights_list[slow:]\n",
    "\n",
    "close_vals_list = close_vals_list[slow:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "departmental-luxembourg",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "peaceful-thumbnail",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_X = df[cols_to_scale]\n",
    "all_y = df['labels']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "adaptive-emission",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_X.head(5)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "devoted-philosophy",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_X.columns.to_list()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "outstanding-browse",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(all_X, all_y, test_size=0.2, shuffle = False)\n",
    "\n",
    "X_train, X_val, y_train, y_val = train_test_split(X_train, y_train, test_size=0.25, shuffle = False) # 0.25 x 0.8 = 0.2\n",
    "\n",
    "#scaling test dataset\n",
    "\n",
    "#X_train_scaled, X_test_scaled, y_train, y_test = train_test_split(all_X_scaled, all_y, test_size=0.2, shuffle = False)\n",
    "\n",
    "#X_train_scaled, X_val_scaled, y_train, y_val = train_test_split(X_train_scaled, y_train, test_size=0.25, shuffle = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "enclosed-patrol",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "express-problem",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_set_len = len(y_test)\n",
    "print(test_set_len)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "modified-danish",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_dates_list = dates_list[-test_set_len:]\n",
    "\n",
    "test_vals_list = close_vals_list[-test_set_len:]\n",
    "\n",
    "test_labels = all_markers[-test_set_len:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "promotional-straight",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "buy_true_test_markers = test_labels.values == 1\n",
    "sell_true_test_markers = test_labels.values == 0\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "renewable-graham",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "asian-sunrise",
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib notebook\n",
    "plt.rcParams['figure.figsize'] = [10, 6]\n",
    "plt.plot(test_dates_list, test_vals_list)\n",
    "plt.plot(test_dates_list, test_vals_list, 'g^', markevery=buy_true_test_markers)\n",
    "plt.plot(test_dates_list, test_vals_list, 'rv', markevery=sell_true_test_markers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "conditional-pakistan",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_len = len(X_train)\n",
    "val_len = len(X_val)\n",
    "test_len = len(X_test)\n",
    "#BUY => 1, SELL => 0, HOLD => 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "alive-physiology",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_weights = weights_list[:train_len]\n",
    "val_weights = weights_list[train_len:train_len + val_len]\n",
    "test_weights = weights_list[train_len + val_len:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "generic-shepherd",
   "metadata": {},
   "outputs": [],
   "source": [
    "#test_weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "useful-riverside",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_test.value_counts()\n",
    "#BUY => 1, SELL => 0, HOLD => 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "charming-satin",
   "metadata": {},
   "outputs": [],
   "source": [
    "d_train = xgb.DMatrix(X_train, label = y_train, weight = train_weights)\n",
    "d_val = xgb.DMatrix(X_val, label = y_val, weight = val_weights)\n",
    "d_test = xgb.DMatrix(X_test, label = y_test, weight = test_weights)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "greater-timing",
   "metadata": {},
   "outputs": [],
   "source": [
    "#val_frame = pd.DataFrame(val_weights)\n",
    "#BUY => 1, SELL => 0, HOLD => 2\n",
    "y_val.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "suitable-beauty",
   "metadata": {},
   "outputs": [],
   "source": [
    "#%reset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "chemical-paint",
   "metadata": {},
   "outputs": [],
   "source": [
    "import yfinance as yf\n",
    "import pandas as pd\n",
    "import datetime\n",
    "import time\n",
    "import matplotlib.pyplot as plt\n",
    "from xgboost import XGBClassifier\n",
    "import xgboost as xgb\n",
    "import numpy as np\n",
    "import pickle\n",
    "import os\n",
    "os.environ[\"PATH\"] += os.pathsep + r'C:\\Program Files\\Graphviz\\bin'\n",
    "yf.pdr_override()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "coordinated-truth",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import recall_score\n",
    "from typing import Tuple\n",
    "\n",
    "def recall(predt: np.ndarray, dtrain: xgb.DMatrix) -> Tuple[str, float]:\n",
    "    ''' Root mean squared log error metric.'''\n",
    "    y = dtrain.get_label()\n",
    "    \n",
    "    inter_vals = pd.DataFrame(y)\n",
    "    vals_all = inter_vals.value_counts()\n",
    "    \n",
    "    #BUY => 1, SELL => 0, HOLD => 2\n",
    "    \n",
    "    #print(predt)\n",
    "    \n",
    "    inter_num_buys = vals_all[1]\n",
    "    inter_num_sells = vals_all[0]\n",
    "    inter_num_holds = vals_all[2]\n",
    "    \n",
    "    least_vals = min(inter_num_buys, inter_num_sells, inter_num_holds)\n",
    "    \n",
    "    imp_factor = 5\n",
    "    \n",
    "    inter_weight_buys = imp_factor * (least_vals / inter_num_buys)\n",
    "    inter_weight_sells = imp_factor * (least_vals / inter_num_sells)\n",
    "    inter_weight_holds = least_vals / inter_num_holds\n",
    "    \n",
    "    \n",
    "    #print(inter_weight_buys)\n",
    "    #print(inter_weight_sells)\n",
    "    #print(inter_weight_holds)\n",
    "    \n",
    "    #predt[predt < -1] = -1 + 1e-6\n",
    "\n",
    "    #print(len(predt))\n",
    "    #print(predt.shape)\n",
    "    \n",
    "    y_pred = [None] * len(predt)\n",
    "    inter_weights = [None] * len(predt)\n",
    "    \n",
    "    for x in range(len(predt)):\n",
    "        \n",
    "        val_0 = predt[x,0]\n",
    "        val_1 = predt[x,1]\n",
    "        val_2 = predt[x,2]\n",
    "        \n",
    "        pos = max(val_0, val_1, val_2)\n",
    "        \n",
    "        if pos == val_0:\n",
    "            y_pred[x] = 0\n",
    "            inter_weights[x] = inter_weight_sells\n",
    "        elif pos == val_1:\n",
    "            y_pred[x] = 1\n",
    "            inter_weights[x] = inter_weight_buys\n",
    "        elif pos == val_2:\n",
    "            y_pred[x] = 2\n",
    "            inter_weights[x] = inter_weight_holds\n",
    "            \n",
    "    #breakpoint\n",
    "    #print(y_pred)\n",
    "    #BUY => 1, SELL => 0, HOLD => 2\n",
    "    \n",
    "    rec_return_sell = recall_score(y, y_pred, labels = [0], average = 'weighted', sample_weight = inter_weights)\n",
    "    rec_return_buy = recall_score(y, y_pred, labels = [1], average = 'weighted', sample_weight = inter_weights)\n",
    "\n",
    "    \n",
    "    #rec_return = (rec_return_sell + rec_return_buy) / 2\n",
    "    #prec_return = (prec_return_sell + prec_return_buy) / 2\n",
    "    \n",
    "    #print(rec_return_sell)\n",
    "    #print(rec_return_buy)\n",
    "    \n",
    "    rec_mult = rec_return_sell * rec_return_buy\n",
    "    \n",
    "    return [('adj_recall', float(rec_mult))]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "reverse-specific",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "spectacular-optimization",
   "metadata": {},
   "outputs": [],
   "source": [
    "d_train = xgb.DMatrix('d_train')\n",
    "d_val = xgb.DMatrix('d_val')\n",
    "d_test = xgb.DMatrix('d_test')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "comic-willow",
   "metadata": {},
   "outputs": [],
   "source": [
    "eval_set = [(d_train, 'train'), (d_val, 'val')]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dominican-hunter",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "assumed-entity",
   "metadata": {},
   "outputs": [],
   "source": [
    "'''xgb.DMatrix.save_binary(d_train, 'd_train')\n",
    "xgb.DMatrix.save_binary(d_val, 'd_val')\n",
    "xgb.DMatrix.save_binary(d_test, 'd_test')'''\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "statistical-papua",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "evals_result = {}\n",
    "\n",
    "params = {'tree_method': 'gpu_hist',\n",
    "            'gpu_id': 0,\n",
    "            'disable_default_eval_metric':True,\n",
    "            'max_depth':29,\n",
    "            'learning_rate':0.5,\n",
    "            #'eval_metric': ['merror','mlogloss'],\n",
    "            'objective': 'multi:softprob',\n",
    "            'min_child_weight':0.5,\n",
    "            'subsample':1,\n",
    "            'num_class': 3,\n",
    "            'gamma':0.1,\n",
    "            'verbosity': 2}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "turkish-welcome",
   "metadata": {},
   "outputs": [],
   "source": [
    "whos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "micro-router",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "aboriginal-bathroom",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "#num_round = 1000\n",
    "bst = xgb.train(params,\n",
    "                d_train,\n",
    "                num_boost_round = 100,\n",
    "                feval = recall,\n",
    "                evals = eval_set,\n",
    "                maximize = True,\n",
    "                evals_result = evals_result,\n",
    "                early_stopping_rounds = 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "rapid-northwest",
   "metadata": {},
   "outputs": [],
   "source": [
    "epochs = len(evals_result['val']['adj_recall'])\n",
    "x_axis = range(0, epochs)\n",
    "\n",
    "\n",
    "plt.rcParams['figure.figsize'] = [10, 6]\n",
    "fig, ax = plt.subplots()\n",
    "ax.plot(x_axis, evals_result['train']['adj_recall'], label='Train')\n",
    "ax.plot(x_axis, evals_result['val']['adj_recall'], label='Val')\n",
    "ax.legend()\n",
    "plt.ylabel('adj_recall')\n",
    "plt.xlabel('n_estimators')\n",
    "plt.title('XGBoost adj_recall')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "accredited-calculator",
   "metadata": {},
   "outputs": [],
   "source": [
    "how_many = bst.best_ntree_limit\n",
    "print(how_many)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "announced-brick",
   "metadata": {},
   "outputs": [],
   "source": [
    "bst.best_iteration\n",
    "\n",
    "#how_many = 5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "empirical-royal",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "bst = xgb.train(params,\n",
    "                d_train,\n",
    "                num_boost_round = how_many,\n",
    "                feval = recall,\n",
    "                maximize = True,\n",
    "                evals = eval_set,\n",
    "                evals_result = evals_result)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "documentary-isaac",
   "metadata": {},
   "outputs": [],
   "source": [
    "preds = bst.predict(d_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "protecting-president",
   "metadata": {},
   "outputs": [],
   "source": [
    "preds_df = pd.DataFrame(preds)\n",
    "\n",
    "preds_df.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "endless-forward",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "latter-ideal",
   "metadata": {},
   "outputs": [],
   "source": [
    "best_preds = np.asarray([np.argmax(line) for line in preds])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "supported-queens",
   "metadata": {},
   "outputs": [],
   "source": [
    "best_preds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dominican-initial",
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_df = pd.DataFrame(best_preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "spanish-latino",
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_df.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "standing-colors",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "olympic-safety",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_test.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "boring-pantyhose",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "killing-turtle",
   "metadata": {},
   "outputs": [],
   "source": [
    "epochs = len(evals_result['val']['adj_recall'])\n",
    "x_axis = range(0, epochs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "found-bacon",
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.rcParams['figure.figsize'] = [10, 6]\n",
    "fig, ax = plt.subplots()\n",
    "ax.plot(x_axis, evals_result['train']['adj_recall'], label='Train')\n",
    "ax.plot(x_axis, evals_result['val']['adj_recall'], label='Val')\n",
    "ax.legend()\n",
    "plt.ylabel('adj_recall')\n",
    "plt.xlabel('n_estimators')\n",
    "plt.title('XGBoost adj_recall')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "celtic-couple",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "enabling-latvia",
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.rcParams['figure.figsize'] = [10, 6]\n",
    "xgb.plot_importance(bst, max_num_features = 50)\n",
    "plt.rcParams['figure.figsize'] = [10, 6]\n",
    "xgb.plot_tree(bst)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "virtual-worry",
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_important = bst.get_score(importance_type='weight')\n",
    "keys = list(feature_important.keys())\n",
    "values = list(feature_important.values())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "capable-railway",
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_dict = dict(zip(keys, values))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "acceptable-bible",
   "metadata": {},
   "outputs": [],
   "source": [
    "my_keys = sorted(feature_dict, key=feature_dict.get, reverse=True)[:29]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "contemporary-acrobat",
   "metadata": {},
   "outputs": [],
   "source": [
    "my_keys.append('labels')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "smoking-faculty",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(my_keys)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "super-married",
   "metadata": {},
   "outputs": [],
   "source": [
    "analyze_df = df[my_keys]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cooked-quantum",
   "metadata": {},
   "outputs": [],
   "source": [
    "analyze_df.to_csv(r'C:\\Users\\Michael\\Desktop\\Python\\Stonks\\Data_Scrape\\file_storage\\analyze.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "mighty-copying",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plt.rcParams['figure.figsize'] = [8, 6]\n",
    "dft = AV.AutoViz(r'C:\\Users\\Michael\\Desktop\\Python\\Stonks\\Data_Scrape\\file_storage\\analyze.csv',\n",
    "                 sep=',', depVar='labels', dfte=analyze_df, header=0, verbose=2,\n",
    "                            lowess=False,chart_format='png',max_rows_analyzed=30000,max_cols_analyzed=30)\n",
    "#BUY => 1, SELL => 0, HOLD => 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "heated-steps",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "compact-differential",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_test.value_counts()\n",
    "#BUY => 1, SELL => 0, HOLD => 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fixed-bangkok",
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_df.value_counts()\n",
    "#BUY => 1, SELL => 0, HOLD => 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "exempt-producer",
   "metadata": {},
   "outputs": [],
   "source": [
    "from imblearn.metrics import classification_report_imbalanced\n",
    "target_names = ['Sell', 'Buy', 'Hold']\n",
    "print(classification_report_imbalanced(y_test.values, pred_df.values, target_names=target_names))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "vulnerable-split",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "conf_mat = confusion_matrix(y_test.values.tolist(), pred_df.values.tolist())\n",
    "conf_mat_norm = confusion_matrix(y_test.values.tolist(), pred_df.values.tolist(), normalize = 'true')\n",
    "print(conf_mat)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "antique-pennsylvania",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import ConfusionMatrixDisplay\n",
    "disp = ConfusionMatrixDisplay(conf_mat_norm, display_labels = target_names)\n",
    "disp.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "graphic-staff",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(pred_df.values[-1])\n",
    "#last class label\n",
    "#plt.ylabel('BUY => 1, SELL => 0, HOLD => 2')\n",
    "test_dates_list[-1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "desperate-accuracy",
   "metadata": {},
   "outputs": [],
   "source": [
    "buy_list_markers = pred_df.index[pred_df[0]==1].tolist()\n",
    "sell_list_markers = pred_df.index[pred_df[0]==0].tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "equivalent-packet",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "%matplotlib notebook\n",
    "plt.rcParams['figure.figsize'] = [10, 6]\n",
    "\n",
    "plt.subplot(2,1,1)\n",
    "\n",
    "plt.plot(test_dates_list, adj_close_list[-len(test_vals_list):])\n",
    "plt.plot(test_dates_list, adj_close_list[-len(test_vals_list):], 'g^', markevery=buy_list_markers)\n",
    "plt.plot(test_dates_list, adj_close_list[-len(test_vals_list):], 'rv', markevery=sell_list_markers)\n",
    "\n",
    "plt.subplot(2,1,2)\n",
    "\n",
    "plt.plot(test_dates_list, adj_close_list[-len(test_vals_list):])\n",
    "plt.plot(test_dates_list, adj_close_list[-len(test_vals_list):], 'g^', markevery=buy_true_test_markers)\n",
    "plt.plot(test_dates_list, adj_close_list[-len(test_vals_list):], 'rv', markevery=sell_true_test_markers)\n",
    "\n",
    "#plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "described-graphic",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "naval-tyler",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_reg = df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "informal-mumbai",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_reg.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "composed-match",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_reg['adj_close'] = close_vals_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fuzzy-algeria",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_reg['adj_close'] = df_reg['adj_close'].shift(30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "legitimate-norway",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_reg = df_reg.tail(-30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "comic-rebecca",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_reg.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "noticed-texas",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_reg.drop(['labels'], axis = 1, inplace = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "governmental-priority",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_reg.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "liable-promise",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_reg['adj_close']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "entertaining-chemical",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_X_reg = df_reg[cols_to_scale]\n",
    "all_y_reg = df_reg['adj_close']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "handled-panama",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "uniform-blogger",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_reg, X_test_reg, y_train_reg, y_test_reg = train_test_split(all_X_reg, all_y_reg, \n",
    "                                                                    test_size=0.2, shuffle = False)\n",
    "\n",
    "X_train_reg, X_val_reg, y_train_reg, y_val_reg = train_test_split(X_train_reg, y_train_reg,\n",
    "                                                                  test_size=0.25, shuffle = False) # 0.25 x 0.8 = 0.2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "metric-colleague",
   "metadata": {},
   "outputs": [],
   "source": [
    "d_train_reg = xgb.DMatrix(X_train_reg, label = y_train_reg)\n",
    "d_val_reg = xgb.DMatrix(X_val_reg, label = y_val_reg)\n",
    "d_test_reg = xgb.DMatrix(X_test_reg, label = y_test_reg)\n",
    "\n",
    "eval_set_reg = [(d_train_reg, 'train'), (d_val_reg, 'val')]\n",
    "\n",
    "evals_result_reg = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "marked-protein",
   "metadata": {},
   "outputs": [],
   "source": [
    "params = {'tree_method': 'gpu_hist',\n",
    "            'gpu_id': 0,\n",
    "            'n_estimators': 500,\n",
    "            'max_depth':25,\n",
    "            'learning_rate':0.1,\n",
    "            'eval_metric': 'mae',\n",
    "            'objective': 'reg:gamma',\n",
    "            'min_child_weight':0.25,\n",
    "            'subsample':1,\n",
    "            'gamma':0.05,\n",
    "            'verbosity': 2}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "large-coral",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "bst = xgb.train(params,\n",
    "                d_train_reg,\n",
    "                num_boost_round = 1000,\n",
    "                evals = eval_set_reg,\n",
    "                evals_result = evals_result_reg,\n",
    "                early_stopping_rounds = 50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "experienced-mayor",
   "metadata": {},
   "outputs": [],
   "source": [
    "epochs = len(evals_result_reg['val']['mae'])\n",
    "x_axis = range(0, epochs)\n",
    "\n",
    "\n",
    "plt.rcParams['figure.figsize'] = [10, 6]\n",
    "fig, ax = plt.subplots()\n",
    "ax.plot(x_axis, evals_result_reg['train']['mae'], label='Train')\n",
    "ax.plot(x_axis, evals_result_reg['val']['mae'], label='Val')\n",
    "\n",
    "ax.legend()\n",
    "plt.ylabel('errors')\n",
    "plt.xlabel('n_estimators')\n",
    "plt.title('XGBoost errors')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "noble-farmer",
   "metadata": {},
   "outputs": [],
   "source": [
    "preds = bst.predict(d_test_reg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bridal-invasion",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(len(preds))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "aggressive-participation",
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib notebook\n",
    "\n",
    "plt.rcParams['figure.figsize'] = [10, 6]\n",
    "\n",
    "plt.plot(test_dates_list[6:], y_test_reg)\n",
    "\n",
    "\n",
    "plt.plot(test_dates_list[6:], preds)\n",
    "\n",
    "plt.legend(['true','pred'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "patent-leadership",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "criminal-metabolism",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}