Encoder-Decorer NQv1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### The Encoder-decoder goal is to build a candle classification dictionary for futher LLM\n",
    "It encodes and decodes <b>normalized candles</b> as there is no goal yet to reproduce price move direction/distance or even correct candle height.\n",
    "This is next step to be implemented with LLM-kind of model, where <i>candles will be considered as alphabet characters</i> and series of candles are words, sentences and chapters for price action poeme :-) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "from torch import nn\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import os.path\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 498,
   "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>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2023-03-13 04:00:00+00:00</th>\n",
       "      <td>12177.25</td>\n",
       "      <td>12179.00</td>\n",
       "      <td>12177.25</td>\n",
       "      <td>12179.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-13 04:00:15+00:00</th>\n",
       "      <td>12179.00</td>\n",
       "      <td>12179.25</td>\n",
       "      <td>12179.00</td>\n",
       "      <td>12179.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-13 04:00:30+00:00</th>\n",
       "      <td>12181.25</td>\n",
       "      <td>12181.25</td>\n",
       "      <td>12180.25</td>\n",
       "      <td>12180.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-13 04:00:45+00:00</th>\n",
       "      <td>12179.50</td>\n",
       "      <td>12179.75</td>\n",
       "      <td>12179.25</td>\n",
       "      <td>12179.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-13 04:01:00+00:00</th>\n",
       "      <td>12180.00</td>\n",
       "      <td>12181.75</td>\n",
       "      <td>12180.00</td>\n",
       "      <td>12180.50</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                               open      high       low     close\n",
       "2023-03-13 04:00:00+00:00  12177.25  12179.00  12177.25  12179.00\n",
       "2023-03-13 04:00:15+00:00  12179.00  12179.25  12179.00  12179.25\n",
       "2023-03-13 04:00:30+00:00  12181.25  12181.25  12180.25  12180.75\n",
       "2023-03-13 04:00:45+00:00  12179.50  12179.75  12179.25  12179.75\n",
       "2023-03-13 04:01:00+00:00  12180.00  12181.75  12180.00  12180.50"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Candle dataset shape: (354167, 4)\n"
     ]
    }
   ],
   "source": [
    "#Import NQ futures candles dataset\n",
    "dataset = './data/nqh23_15sec.csv'\n",
    "tick_data = './data/NQ 06-23.Last.csv'\n",
    "if os.path.isfile(dataset):\n",
    "    df1 = pd.read_csv(dataset)\n",
    "\n",
    "else:\n",
    "    header = ('date','last')\n",
    "    df = pd.read_csv(tick_data, delimiter=';', usecols=[0,1], header=None, names=header)\n",
    "    df['date'] = pd.to_datetime(df['date'], format='%Y%m%d %H%M%S %f', utc = True)\n",
    "    df1 = df.resample(\"15s\", on = 'date',).ohlc()['last']\n",
    "    df1.dropna(inplace=True)\n",
    "    df1.to_csv(dataset)\n",
    "\n",
    "\n",
    "df1.set_index('date',inplace=True)\n",
    "df1.index.name = None\n",
    "display(df1.head())\n",
    "print(f\"Candle dataset shape: {df1.shape}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# My own Normalisation function. the only reason I use them is I can easie reproduce them in TradingView Pinescript, \n",
    "# while e.g Pytorch nn.instanceNorm1d, gives sllughtly different result. But technically there is no differece which one to use here.\n",
    "def MinMaxNorm(x: torch.Tensor, dim = None, shift = False):\n",
    "\n",
    "    r'''MinMax Normalization function, normalize values in (0,1) range.\n",
    "\n",
    "    dim - dimentions setting, \n",
    "    where None - for global normalization, 1 - for per instance normalizaton and 0 - per feature normalization\n",
    "    shif = True shifts all values agains candle Open'''\n",
    "\n",
    "    if dim == None:\n",
    "        return (x - x.min())/(x.max() - x.min())\n",
    "    elif dim == 1:\n",
    "        if shift:\n",
    "            norm_x = (torch.nan_to_num((x - x.min(dim).values.view(-1,1))/(x.max(dim).values - x.min(dim).values).view(-1,1),0))\n",
    "            return  norm_x - norm_x[:,0].view(-1,1) \n",
    "        else:\n",
    "            return torch.nan_to_num((x - x.min(dim).values.view(-1,1))/(x.max(dim).values - x.min(dim).values).view(-1,1),0)\n",
    "        \n",
    "    elif dim == 0:\n",
    "        return (x - x.min(dim).values)/(x.max(dim).values - x.min(dim).values)\n",
    "\n",
    "    \n",
    "\n",
    "def StandardNorm(x: torch.Tensor, dim = None):\n",
    "\n",
    "    r'''Standard Normalization function, normalize values in (0,1) range.\n",
    "\n",
    "    dim - dimentions setting, \n",
    "    where None - for global normalization, 1 - for per instance normalizaton and 0 - per feature normalization'''\n",
    "\n",
    "    if dim == None:\n",
    "        return (x - x.mean())/(x.std())\n",
    "    elif dim ==1:\n",
    "        return (x - x.mean(dim).reshape(-1,1))/(x.std(dim).reshape(-1,1))\n",
    "    else:\n",
    "        return (x - x.mean(dim))/(x.std(dim))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 499,
   "metadata": {},
   "outputs": [],
   "source": [
    "# in addition to Loss control< good to use a candle decode accuracyEncode-Decode accuracy calculation function\n",
    "\n",
    "def accuracy(actual:torch.Tensor, pred:torch.Tensor, precision):\n",
    "    return (torch.isclose(actual, pred, precision, precision).sum().item()/ (pred.shape[0] * pred.shape[1]))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 500,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[12177.2500, 12179.0000, 12177.2500, 12179.0000],\n",
       "        [12179.0000, 12179.2500, 12179.0000, 12179.2500],\n",
       "        [12181.2500, 12181.2500, 12180.2500, 12180.7500],\n",
       "        ...,\n",
       "        [14864.5000, 14866.0000, 14864.5000, 14866.0000],\n",
       "        [14866.7500, 14867.0000, 14866.5000, 14866.5000],\n",
       "        [14867.2500, 14867.2500, 14865.0000, 14866.2500]])"
      ]
     },
     "execution_count": 500,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#pick 10K candles as traning sample and conver to torch tensor\n",
    "\n",
    "candles_sample = torch.tensor(df1.iloc[:].values,dtype=torch.float)\n",
    "candles_sample\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 584,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function which generate series of dots, reflecting price flow while a candle shapes.\n",
    "# The assumption is that for GREEN candle the price moves as open->low->high->close \n",
    "# and for RED the price moves as open->high->low->close \n",
    "# Function returns a tensor with 30 priceing dots\n",
    "\n",
    "def CandlePriceFlow(sample : torch.Tensor):\n",
    "    cdl = sample.clone()\n",
    "    step1 = torch.where(cdl[:,0] <= cdl[:,3], cdl[:,2], cdl[:,1])\n",
    "    step2 = torch.where(cdl[:,0] <= cdl[:,3], cdl[:,1], cdl[:,2])\n",
    "    #converting high-low sequence for green candles\n",
    "    cdl[:,1] = step1\n",
    "    cdl[:,2] = step2\n",
    "    \n",
    "    # create three lines for price flow: open-low, low-high, high-close (this is for green candle, for red is similar)\n",
    "\n",
    "    y = torch.concatenate([cdl[:,:2],cdl[:,1:3],cdl[:,2:4]],dim=1)\n",
    "    x = torch.zeros((1,6))\n",
    "    for i in range(3):\n",
    "        x[0,i*2:i*2+2] = torch.tensor([i*10,(i+1)*10])\n",
    "\n",
    "    #create empty a and b variables tensors for linear equation a * x + b for for price move sloap.\n",
    "\n",
    "    a = torch.zeros((len(y),3))\n",
    "    b = torch.zeros((len(y),3))\n",
    "    # calculate three sets of a,b variblesfor three price move sloap\n",
    "\n",
    "    for i in range(3):\n",
    "        a[:,i] = (y[:,i*2+1]-y[:,i*2]) * (x[:,i*2+1]-x[:,i*2]) ** -1\n",
    "        b[:,i] = (x[:,i*2+1]*y[:,i*2] -  y[:,i*2+1]*x[:,i*2]) * (x[:,i*2+1]-x[:,i*2]) ** -1\n",
    "    \n",
    "    x = torch.arange(0,30,dtype=torch.float).reshape(3,-1)\n",
    "    \n",
    "    # calculate price  numbers, 10 for for each of price move sloap and merge in a single tensor with size [number_of_candles,30]\n",
    "    \n",
    "    return  torch.concat([a[:,0].view(-1,1) @ x[0].view(1,-1) + b[:,0].view(-1,1),\n",
    "                a[:,1].view(-1,1) @ x[1].view(1,-1) + b[:,1].view(-1,1),\n",
    "                a[:,2].view(-1,1) @ x[2].view(1,-1) + b[:,2].view(-1,1)],dim=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 501,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[0.0000, 1.0000, 0.0000, 1.0000],\n",
       "        [0.0000, 1.0000, 0.0000, 1.0000],\n",
       "        [1.0000, 1.0000, 0.0000, 0.5000],\n",
       "        ...,\n",
       "        [0.0000, 1.0000, 0.0000, 1.0000],\n",
       "        [0.5000, 1.0000, 0.0000, 0.0000],\n",
       "        [1.0000, 1.0000, 0.0000, 0.5556]])"
      ]
     },
     "execution_count": 501,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#normalize traning set of candles as encode-decode sample\n",
    "\n",
    "candles_norm = MinMaxNorm((candles_sample),1)\n",
    "candles_norm\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 502,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Enc-Dec NN model definition. This is rather simple Linear with one hidden layer and Softmax activation \n",
    "# to get candles class codes (set of 10 candle features probabilities)\n",
    "# Trained Encoder half of the model can be used later on for quick canldes classification.\n",
    "# self.label is out put of the ecoder, which is indeed a candle class code\n",
    "\n",
    "class CandleClassifier(nn.Module):\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "       \n",
    "    #encoding layers\n",
    "        self.seq1 = nn.Sequential(\n",
    "            nn.Linear(4,16),\n",
    "            nn.ReLU(),\n",
    "            nn.Linear(16,2),\n",
    "            nn.ReLU()\n",
    "            )\n",
    "\n",
    "    #decoding layers    \n",
    "        self.seq2 = nn.Sequential(\n",
    "            nn.Linear(2,16),\n",
    "            nn.ReLU(),\n",
    "            nn.Linear(16,4),\n",
    "            nn.Sigmoid()\n",
    "        )\n",
    "        \n",
    "\n",
    "    def forward(self, x):\n",
    "        self.label = (self.seq1(x))\n",
    "        return self.seq2(self.label)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 505,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 0 | Loss: 0.7058794498443604 | Accuracy: 0.02%\n",
      "\n",
      "Epoch: 883 | Loss: 0.16059400141239166 | Accuracy: 49.62%\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Traning accuracy')"
      ]
     },
     "execution_count": 505,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x300 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Encode-Decode Traning loop\n",
    "\n",
    "torch.manual_seed(2345163)\n",
    "loss_plt = []\n",
    "acc_plt = []\n",
    "model0 = CandleClassifier()\n",
    "\n",
    "epochs = 884\n",
    "model0 = model0.to(\"cpu\")\n",
    "candles_norm = candles_norm.to('cpu')\n",
    "loss_fn =  nn.BCELoss()\n",
    "optimizer = torch.optim.Adam(model0.parameters(), lr = 0.04, weight_decay=1e-4) # Adam optimizer was choosen insted of SGD due convergency issues with SGD\n",
    "\n",
    "      \n",
    "for epoch in range(epochs):\n",
    "\n",
    "    #set train mode\n",
    "    model0.train()\n",
    "\n",
    "    #forward pass\n",
    "    out = model0(candles_norm)\n",
    "    \n",
    "    #loss and accuracy calculation\n",
    "    loss = loss_fn(out, candles_norm)\n",
    "    acc = accuracy(out, candles_norm, 1e-3)\n",
    "    \n",
    "    optimizer.zero_grad()\n",
    "    loss.backward()\n",
    "    optimizer.step()\n",
    "    \n",
    "    #calculate and collect loss and accucary values through the traning fot further plot and print\n",
    "    if epoch % 1 == 0:\n",
    "        loss_plt.append(loss.item())\n",
    "        acc_plt.append(acc)\n",
    "    if epoch % (epochs/10) == 0:\n",
    "        print(f\"Epoch: {epoch} | Loss: {loss} | Accuracy: %.2f\" % (acc*100)+\"%\")\n",
    "    #Break the traning loop in case the accuracy 99% is achieved    \n",
    "    # if acc > 0.50 or loss < 0.37146:\n",
    "    #     break\n",
    "\n",
    "\n",
    "#adding final Loss and accucacy    \n",
    "loss_plt.append(loss.item())\n",
    "acc_plt.append(acc)\n",
    "print(f\"\\nEpoch: {epoch} | Loss: {loss} | Accuracy: %.2f\" % (acc*100)+\"%\")\n",
    "\n",
    "#Plot loss function and accuracy change through the traning\n",
    "plt.figure(figsize=(12,3))\n",
    "plt.subplot(121)\n",
    "plt.plot(loss_plt)\n",
    "plt.scatter(x=loss_plt.index(min(loss_plt)),y=min(loss_plt),color='red', marker='.')\n",
    "plt.title('Traning loss function')\n",
    "\n",
    "plt.subplot(122)\n",
    "plt.plot(acc_plt, color='orange')\n",
    "plt.title('Traning accuracy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 506,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(883, 0.16059400141239166, 800, 0.5028538796669367)"
      ]
     },
     "execution_count": 506,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loss_plt.index(min(loss_plt)),min(loss_plt), acc_plt.index(max(acc_plt)),max(acc_plt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 281,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1600x800 with 50 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot example of 50 candles labels(classes) (kind of candle class code. )\n",
    "#These candles are plotted in the next notebook cell\n",
    "\n",
    "plt.figure(figsize=(16,8))\n",
    "plt.rcParams.update({'font.size': 6})\n",
    "#y = torch.tensor(list(enumerate(model0.label[1000:1050].cpu().detach().numpy())))[:,1]\n",
    "for i in range(0,50,1):\n",
    "\n",
    "#    print(f\"Candle {i}: {y[i].item(): .3f} | Candle {i+1}: {y[i+1].item(): .3f} | Candle {i+2}: {y[i+2].item(): .3f} | Candle {i+3}: {y[i+3].item(): .3f} | Candle {i+4}: {y[i+4].item(): .3f} \")\n",
    "    \n",
    "    plt.subplot(5,10,i+1)\n",
    "    plt.title(f'Candle #{i}')\n",
    "    x = torch.tensor(list(enumerate(model0.label[1000+i].cpu().detach().numpy())))[:,0]\n",
    "    y = torch.round(torch.tensor(list(enumerate(model0.label[1000+i].cpu().detach().numpy())))[:,1])\n",
    "    plt.bar(x,y)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Plot candle price flow\n",
    "plt.figure(figsize=(16,10))\n",
    "for i in range(50):\n",
    "    plt.subplot(5,10,i+1)\n",
    "    plt.title(f'Candle #{i}')\n",
    "    plt.plot(candles_norm[1000+i].cpu().detach().numpy())\n",
    "    plt.plot(out[1000+i].cpu().detach().numpy(), alpha=0.7)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "##Function to plot candles from the dataset generated based on prediction\n",
    "def plot_sample(sample):\n",
    "    import plotly.graph_objects as go\n",
    "\n",
    "    cdl_data = pd.DataFrame(sample, columns=('open','high','low','close') )\n",
    "\n",
    "    fig = go.Figure(data=[go.Candlestick(x=cdl_data.index,\n",
    "                    open=cdl_data['open'], high=cdl_data['high'],\n",
    "                    low=cdl_data['low'], close=cdl_data['close'])\n",
    "                        ])\n",
    "\n",
    "    # fig.add_trace(\n",
    "    #     go.Scatter(\n",
    "    #         x=cdl_data.index,\n",
    "    #         y=cdl_data['high'],\n",
    "    #         mode= \"text\",\n",
    "            \n",
    "    #         text = cdl_data['class'],\n",
    "    #         textposition = 'top center',\n",
    "    #         textfont = {'size':8}))\n",
    "\n",
    "    fig.update_layout(xaxis_rangeslider_visible=False)\n",
    "    fig.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 507,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "plotlyServerURL": "https://plot.ly"
       },
       "data": [
        {
         "close": [
          0.22417941689491272,
          0.5243338942527771,
          0.32236236333847046,
          0.9816672801971436,
          0.32275885343551636,
          0.5733383893966675,
          0.8963143825531006,
          0.0048304712399840355,
          0.13793642818927765,
          0.40274229645729065,
          0.07612400501966476,
          0.3232440948486328,
          0.32214200496673584,
          0.6777097582817078,
          0.00483598280698061,
          0.9634014368057251,
          0.1642286330461502,
          0.5620012283325195,
          0.004812444094568491,
          0.86729496717453,
          0.6090378165245056,
          0.5469033718109131,
          0.0048355949111282825,
          0.6781290769577026,
          0.24730105698108673,
          0.9816672801971436,
          0.025164321064949036,
          0.02929430641233921,
          0.8312090039253235,
          0.4220602512359619,
          0.004820785019546747,
          0.3908326029777527,
          0.6505944132804871,
          0.9670943021774292,
          0.1598832756280899,
          0.004827162250876427,
          0.004836230538785458,
          0.5807435512542725,
          0.20754767954349518,
          0.7727978229522705,
          0.864797830581665,
          0.9695714116096497,
          0.7295553088188171,
          0.009034266695380211,
          0.9576278924942017,
          0.9318862557411194,
          0.3001019358634949,
          0.5807493925094604,
          0.36956819891929626,
          0.50179123878479
         ],
         "high": [
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          0.9999998807907104,
          1,
          1,
          1,
          1,
          1,
          0.9999942779541016,
          0.999634861946106,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          0.9998316764831543,
          1,
          1,
          1,
          0.9999979734420776,
          1,
          1,
          0.9999810457229614,
          1,
          1,
          1,
          0.999958872795105,
          1,
          0.9999513626098633,
          1,
          0.9999816417694092,
          1,
          1,
          0.9999997615814209,
          1,
          1,
          0.9999973773956299,
          0.9999717473983765,
          0.9999972581863403
         ],
         "low": [
          3.3338099569846236e-7,
          5.426934990282462e-7,
          0.0000013495750863512512,
          2.8824410946981516e-7,
          0.000004444464138941839,
          0.000006837185992480954,
          0.000003372317678440595,
          0.000004852084202866536,
          7.159598567341163e-7,
          4.967569680047745e-7,
          5.645347300742287e-7,
          0.000019074959709541872,
          6.960324299143394e-7,
          1.853511122362761e-7,
          0.0000011286130074950051,
          9.354346843792882e-7,
          0.0000010661816531865043,
          0.00006403319275705144,
          0.0005810288130305707,
          1.3118474306850203e-8,
          4.512489581998125e-8,
          0.0000016841038359416416,
          0.0000012503046491474379,
          6.54254392884468e-7,
          0.000005664147920469986,
          2.8824410946981516e-7,
          0.0005377070046961308,
          7.29840053281805e-7,
          0.000001279039565815765,
          0.000010481662684469484,
          0.0000633489980828017,
          0.000003122843281744281,
          3.651357260991972e-8,
          0.000009543326086713932,
          0.000021620726329274476,
          0.00001167906157206744,
          0.0000010598324706734275,
          0.00006934690463822335,
          6.369197080857703e-7,
          0.000027381196559872478,
          0.0000026205661924905144,
          0.000009200628483085893,
          3.3396142384845007e-7,
          8.065420047387306e-7,
          0.000007426562660839409,
          4.3251578318859174e-8,
          6.256319124986476e-7,
          0.000047632336645619944,
          0.00017660934827290475,
          0.00005753911318606697
         ],
         "open": [
          0.9483387470245361,
          0.8114479184150696,
          0.8175141215324402,
          0.5349344611167908,
          0.6560323238372803,
          0.37388017773628235,
          0.2522886395454407,
          0.8315927386283875,
          0.9443410634994507,
          0.8729430437088013,
          0.9644766449928284,
          0.4017849564552307,
          0.8780403733253479,
          0.8450345993041992,
          0.9445950984954834,
          0.3787113130092621,
          0.915916383266449,
          0.09720787405967712,
          0.07803346961736679,
          0.9515343904495239,
          0.9504320025444031,
          0.6398717761039734,
          0.9398615956306458,
          0.6883768439292908,
          0.6853293180465698,
          0.5349344611167908,
          0.08223377168178558,
          0.9579901099205017,
          0.44397416710853577,
          0.42085200548171997,
          0.35754892230033875,
          0.6577654480934143,
          0.9506489634513855,
          0.005900971125811338,
          0.5592240691184998,
          0.7007015943527222,
          0.9473254084587097,
          0.025527888908982277,
          0.9260757565498352,
          0.008987662382423878,
          0.30646613240242004,
          0.0058150505647063255,
          0.7446042895317078,
          0.9566512703895569,
          0.1198757141828537,
          0.8702309727668762,
          0.8942530751228333,
          0.12690798938274384,
          0.07869259268045425,
          0.1416601836681366
         ],
         "type": "candlestick",
         "x": [
          0,
          1,
          2,
          3,
          4,
          5,
          6,
          7,
          8,
          9,
          10,
          11,
          12,
          13,
          14,
          15,
          16,
          17,
          18,
          19,
          20,
          21,
          22,
          23,
          24,
          25,
          26,
          27,
          28,
          29,
          30,
          31,
          32,
          33,
          34,
          35,
          36,
          37,
          38,
          39,
          40,
          41,
          42,
          43,
          44,
          45,
          46,
          47,
          48,
          49
         ]
        }
       ],
       "layout": {
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "fillpattern": {
             "fillmode": "overlay",
             "size": 10,
             "solidity": 0.2
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "autotypenumbers": "strict",
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        },
        "xaxis": {
         "rangeslider": {
          "visible": false
         }
        }
       }
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "plotlyServerURL": "https://plot.ly"
       },
       "data": [
        {
         "close": [
          12159.25,
          12159.25,
          12156.75,
          12158,
          12157,
          12158.25,
          12162,
          12156.75,
          12152,
          12150.5,
          12145.5,
          12145,
          12142.75,
          12142.5,
          12135.25,
          12138.75,
          12133.75,
          12135,
          12134.75,
          12135.5,
          12133,
          12132,
          12126,
          12126,
          12123,
          12127.75,
          12127.25,
          12119.5,
          12121.25,
          12120.5,
          12116.75,
          12115.75,
          12113.5,
          12118.5,
          12117.5,
          12116.5,
          12108,
          12112.75,
          12110.25,
          12115.5,
          12115.25,
          12122.25,
          12122.25,
          12112.75,
          12118.5,
          12118.25,
          12112.5,
          12112,
          12112.25,
          12114
         ],
         "high": [
          12162.75,
          12162,
          12158.75,
          12158,
          12159.5,
          12160.75,
          12162.75,
          12162.25,
          12156.5,
          12153,
          12150,
          12147,
          12144.75,
          12143.25,
          12141,
          12139,
          12137.75,
          12136,
          12139,
          12136,
          12134.5,
          12134.75,
          12133.5,
          12127.75,
          12127.5,
          12127.75,
          12131,
          12126.75,
          12122,
          12122.5,
          12120.5,
          12118,
          12115.75,
          12119,
          12120.75,
          12119.75,
          12117.5,
          12115.25,
          12113.25,
          12117.5,
          12115.5,
          12122.75,
          12123.5,
          12123.75,
          12119,
          12119.25,
          12117.5,
          12112.75,
          12115.5,
          12116.25
         ],
         "low": [
          12158.25,
          12156.25,
          12155.75,
          12155,
          12155.75,
          12155,
          12157.5,
          12156.75,
          12151.25,
          12148.75,
          12145,
          12144,
          12141.75,
          12141,
          12135.25,
          12134.5,
          12133,
          12133.75,
          12134.75,
          12133,
          12130.75,
          12128.75,
          12126,
          12122.5,
          12121.5,
          12122.75,
          12127,
          12119,
          12118,
          12119,
          12116.75,
          12114.25,
          12109.5,
          12112.5,
          12117,
          12116.5,
          12108,
          12109.25,
          12109.5,
          12109,
          12114,
          12115.75,
          12119,
          12112.5,
          12111.5,
          12110,
          12110.25,
          12111,
          12110.25,
          12111.75
         ],
         "open": [
          12162.75,
          12160.5,
          12158,
          12156.5,
          12158,
          12157.25,
          12159,
          12160.75,
          12156.25,
          12152.25,
          12150,
          12145.25,
          12144.25,
          12142.75,
          12140.5,
          12136.25,
          12137.25,
          12134,
          12135.25,
          12136,
          12134.5,
          12132.25,
          12132.75,
          12125.75,
          12125.25,
          12125.25,
          12127.5,
          12126.5,
          12119.75,
          12120.5,
          12118.25,
          12116.5,
          12115.75,
          12112.5,
          12119,
          12118.5,
          12116.75,
          12109.5,
          12113,
          12109,
          12114.5,
          12115.75,
          12122,
          12123.25,
          12112.5,
          12117.5,
          12116.5,
          12111.25,
          12110.75,
          12112.5
         ],
         "type": "candlestick",
         "x": [
          0,
          1,
          2,
          3,
          4,
          5,
          6,
          7,
          8,
          9,
          10,
          11,
          12,
          13,
          14,
          15,
          16,
          17,
          18,
          19,
          20,
          21,
          22,
          23,
          24,
          25,
          26,
          27,
          28,
          29,
          30,
          31,
          32,
          33,
          34,
          35,
          36,
          37,
          38,
          39,
          40,
          41,
          42,
          43,
          44,
          45,
          46,
          47,
          48,
          49
         ]
        }
       ],
       "layout": {
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "fillpattern": {
             "fillmode": "overlay",
             "size": 10,
             "solidity": 0.2
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "autotypenumbers": "strict",
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        },
        "xaxis": {
         "rangeslider": {
          "visible": false
         }
        }
       }
      }
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot candlestick charts\n",
    "plot_sample(out[1000:1050].detach().numpy())\n",
    "plot_sample(candles_sample[1000:1050])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 508,
   "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>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>codes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2023-04-20 22:44:15+00:00</th>\n",
       "      <td>13081.00</td>\n",
       "      <td>13081.00</td>\n",
       "      <td>13081.00</td>\n",
       "      <td>13081.00</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-04-19 23:09:45+00:00</th>\n",
       "      <td>13147.00</td>\n",
       "      <td>13147.50</td>\n",
       "      <td>13147.00</td>\n",
       "      <td>13147.00</td>\n",
       "      <td>12.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-06 00:18:00+00:00</th>\n",
       "      <td>14561.00</td>\n",
       "      <td>14561.75</td>\n",
       "      <td>14560.75</td>\n",
       "      <td>14560.75</td>\n",
       "      <td>13.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-05-15 10:25:00+00:00</th>\n",
       "      <td>13432.50</td>\n",
       "      <td>13433.25</td>\n",
       "      <td>13432.50</td>\n",
       "      <td>13432.75</td>\n",
       "      <td>21.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-04-11 14:20:45+00:00</th>\n",
       "      <td>13085.75</td>\n",
       "      <td>13089.75</td>\n",
       "      <td>13085.25</td>\n",
       "      <td>13086.25</td>\n",
       "      <td>22.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-04-10 00:04:00+00:00</th>\n",
       "      <td>13157.75</td>\n",
       "      <td>13158.25</td>\n",
       "      <td>13157.50</td>\n",
       "      <td>13157.50</td>\n",
       "      <td>23.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-17 04:05:00+00:00</th>\n",
       "      <td>12721.00</td>\n",
       "      <td>12721.25</td>\n",
       "      <td>12720.75</td>\n",
       "      <td>12720.75</td>\n",
       "      <td>24.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-05-03 22:09:30+00:00</th>\n",
       "      <td>13063.25</td>\n",
       "      <td>13064.00</td>\n",
       "      <td>13061.75</td>\n",
       "      <td>13061.75</td>\n",
       "      <td>25.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-02 03:31:00+00:00</th>\n",
       "      <td>14503.00</td>\n",
       "      <td>14504.75</td>\n",
       "      <td>14503.00</td>\n",
       "      <td>14503.75</td>\n",
       "      <td>31.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-05-30 14:10:00+00:00</th>\n",
       "      <td>14438.25</td>\n",
       "      <td>14442.75</td>\n",
       "      <td>14437.25</td>\n",
       "      <td>14438.75</td>\n",
       "      <td>32.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-23 00:28:45+00:00</th>\n",
       "      <td>12720.00</td>\n",
       "      <td>12720.50</td>\n",
       "      <td>12719.75</td>\n",
       "      <td>12720.00</td>\n",
       "      <td>33.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-04-24 12:36:30+00:00</th>\n",
       "      <td>13075.75</td>\n",
       "      <td>13076.75</td>\n",
       "      <td>13074.75</td>\n",
       "      <td>13075.00</td>\n",
       "      <td>34.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-04-03 16:17:00+00:00</th>\n",
       "      <td>13204.75</td>\n",
       "      <td>13205.75</td>\n",
       "      <td>13202.00</td>\n",
       "      <td>13202.75</td>\n",
       "      <td>35.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-05-25 14:53:15+00:00</th>\n",
       "      <td>13923.25</td>\n",
       "      <td>13923.50</td>\n",
       "      <td>13918.50</td>\n",
       "      <td>13918.75</td>\n",
       "      <td>36.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-04-11 10:44:00+00:00</th>\n",
       "      <td>13186.00</td>\n",
       "      <td>13186.00</td>\n",
       "      <td>13185.75</td>\n",
       "      <td>13185.75</td>\n",
       "      <td>37.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-05-09 06:59:00+00:00</th>\n",
       "      <td>13323.50</td>\n",
       "      <td>13325.50</td>\n",
       "      <td>13323.50</td>\n",
       "      <td>13325.00</td>\n",
       "      <td>40.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-04-13 14:16:45+00:00</th>\n",
       "      <td>13124.50</td>\n",
       "      <td>13129.50</td>\n",
       "      <td>13123.75</td>\n",
       "      <td>13127.50</td>\n",
       "      <td>41.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-23 16:22:45+00:00</th>\n",
       "      <td>12969.00</td>\n",
       "      <td>12973.50</td>\n",
       "      <td>12967.50</td>\n",
       "      <td>12970.50</td>\n",
       "      <td>42.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-05-15 19:35:15+00:00</th>\n",
       "      <td>13458.75</td>\n",
       "      <td>13460.25</td>\n",
       "      <td>13457.25</td>\n",
       "      <td>13458.75</td>\n",
       "      <td>43.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-05-30 16:47:00+00:00</th>\n",
       "      <td>14413.50</td>\n",
       "      <td>14415.00</td>\n",
       "      <td>14410.50</td>\n",
       "      <td>14412.00</td>\n",
       "      <td>44.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-21 15:08:45+00:00</th>\n",
       "      <td>12742.00</td>\n",
       "      <td>12742.75</td>\n",
       "      <td>12739.50</td>\n",
       "      <td>12740.75</td>\n",
       "      <td>45.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-04-28 06:04:30+00:00</th>\n",
       "      <td>13216.25</td>\n",
       "      <td>13216.25</td>\n",
       "      <td>13215.25</td>\n",
       "      <td>13215.50</td>\n",
       "      <td>46.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-05-11 18:22:15+00:00</th>\n",
       "      <td>13424.00</td>\n",
       "      <td>13428.00</td>\n",
       "      <td>13424.00</td>\n",
       "      <td>13427.75</td>\n",
       "      <td>50.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-05-29 04:59:45+00:00</th>\n",
       "      <td>14394.25</td>\n",
       "      <td>14395.25</td>\n",
       "      <td>14394.00</td>\n",
       "      <td>14395.00</td>\n",
       "      <td>51.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-28 05:27:00+00:00</th>\n",
       "      <td>12796.25</td>\n",
       "      <td>12796.75</td>\n",
       "      <td>12796.00</td>\n",
       "      <td>12796.50</td>\n",
       "      <td>52.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-01 18:28:00+00:00</th>\n",
       "      <td>14493.75</td>\n",
       "      <td>14494.75</td>\n",
       "      <td>14493.00</td>\n",
       "      <td>14494.00</td>\n",
       "      <td>53.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-13 19:22:15+00:00</th>\n",
       "      <td>12124.50</td>\n",
       "      <td>12126.00</td>\n",
       "      <td>12120.25</td>\n",
       "      <td>12123.25</td>\n",
       "      <td>54.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-04-03 14:43:30+00:00</th>\n",
       "      <td>13238.75</td>\n",
       "      <td>13239.25</td>\n",
       "      <td>13236.00</td>\n",
       "      <td>13237.25</td>\n",
       "      <td>55.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-05-04 01:44:45+00:00</th>\n",
       "      <td>13128.50</td>\n",
       "      <td>13128.50</td>\n",
       "      <td>13127.75</td>\n",
       "      <td>13128.00</td>\n",
       "      <td>56.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-13 00:48:45+00:00</th>\n",
       "      <td>14816.50</td>\n",
       "      <td>14817.00</td>\n",
       "      <td>14816.50</td>\n",
       "      <td>14817.00</td>\n",
       "      <td>60.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-06 01:01:30+00:00</th>\n",
       "      <td>14573.75</td>\n",
       "      <td>14575.00</td>\n",
       "      <td>14573.25</td>\n",
       "      <td>14575.00</td>\n",
       "      <td>61.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-05-03 17:25:15+00:00</th>\n",
       "      <td>13223.75</td>\n",
       "      <td>13225.00</td>\n",
       "      <td>13223.25</td>\n",
       "      <td>13224.75</td>\n",
       "      <td>62.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-09 11:57:15+00:00</th>\n",
       "      <td>14516.50</td>\n",
       "      <td>14517.00</td>\n",
       "      <td>14515.75</td>\n",
       "      <td>14516.75</td>\n",
       "      <td>63.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-23 07:25:00+00:00</th>\n",
       "      <td>12812.25</td>\n",
       "      <td>12812.50</td>\n",
       "      <td>12811.50</td>\n",
       "      <td>12812.25</td>\n",
       "      <td>64.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-08 18:08:15+00:00</th>\n",
       "      <td>14443.50</td>\n",
       "      <td>14443.50</td>\n",
       "      <td>14441.25</td>\n",
       "      <td>14442.75</td>\n",
       "      <td>65.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-05-08 19:29:00+00:00</th>\n",
       "      <td>13345.25</td>\n",
       "      <td>13345.25</td>\n",
       "      <td>13340.25</td>\n",
       "      <td>13343.25</td>\n",
       "      <td>66.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-29 04:24:15+00:00</th>\n",
       "      <td>12778.00</td>\n",
       "      <td>12778.25</td>\n",
       "      <td>12777.75</td>\n",
       "      <td>12778.25</td>\n",
       "      <td>72.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-04-12 05:46:30+00:00</th>\n",
       "      <td>13076.25</td>\n",
       "      <td>13076.50</td>\n",
       "      <td>13075.75</td>\n",
       "      <td>13076.50</td>\n",
       "      <td>73.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-05-26 06:06:15+00:00</th>\n",
       "      <td>13963.00</td>\n",
       "      <td>13963.25</td>\n",
       "      <td>13962.25</td>\n",
       "      <td>13963.25</td>\n",
       "      <td>74.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-04-10 11:00:45+00:00</th>\n",
       "      <td>13152.00</td>\n",
       "      <td>13152.00</td>\n",
       "      <td>13151.00</td>\n",
       "      <td>13151.75</td>\n",
       "      <td>75.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-05-29 07:14:00+00:00</th>\n",
       "      <td>14409.50</td>\n",
       "      <td>14409.75</td>\n",
       "      <td>14408.25</td>\n",
       "      <td>14409.75</td>\n",
       "      <td>84.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-05-28 23:42:45+00:00</th>\n",
       "      <td>14419.00</td>\n",
       "      <td>14419.00</td>\n",
       "      <td>14418.75</td>\n",
       "      <td>14419.00</td>\n",
       "      <td>85.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                               open      high       low     close  codes\n",
       "2023-04-20 22:44:15+00:00  13081.00  13081.00  13081.00  13081.00    0.0\n",
       "2023-04-19 23:09:45+00:00  13147.00  13147.50  13147.00  13147.00   12.0\n",
       "2023-06-06 00:18:00+00:00  14561.00  14561.75  14560.75  14560.75   13.0\n",
       "2023-05-15 10:25:00+00:00  13432.50  13433.25  13432.50  13432.75   21.0\n",
       "2023-04-11 14:20:45+00:00  13085.75  13089.75  13085.25  13086.25   22.0\n",
       "2023-04-10 00:04:00+00:00  13157.75  13158.25  13157.50  13157.50   23.0\n",
       "2023-03-17 04:05:00+00:00  12721.00  12721.25  12720.75  12720.75   24.0\n",
       "2023-05-03 22:09:30+00:00  13063.25  13064.00  13061.75  13061.75   25.0\n",
       "2023-06-02 03:31:00+00:00  14503.00  14504.75  14503.00  14503.75   31.0\n",
       "2023-05-30 14:10:00+00:00  14438.25  14442.75  14437.25  14438.75   32.0\n",
       "2023-03-23 00:28:45+00:00  12720.00  12720.50  12719.75  12720.00   33.0\n",
       "2023-04-24 12:36:30+00:00  13075.75  13076.75  13074.75  13075.00   34.0\n",
       "2023-04-03 16:17:00+00:00  13204.75  13205.75  13202.00  13202.75   35.0\n",
       "2023-05-25 14:53:15+00:00  13923.25  13923.50  13918.50  13918.75   36.0\n",
       "2023-04-11 10:44:00+00:00  13186.00  13186.00  13185.75  13185.75   37.0\n",
       "2023-05-09 06:59:00+00:00  13323.50  13325.50  13323.50  13325.00   40.0\n",
       "2023-04-13 14:16:45+00:00  13124.50  13129.50  13123.75  13127.50   41.0\n",
       "2023-03-23 16:22:45+00:00  12969.00  12973.50  12967.50  12970.50   42.0\n",
       "2023-05-15 19:35:15+00:00  13458.75  13460.25  13457.25  13458.75   43.0\n",
       "2023-05-30 16:47:00+00:00  14413.50  14415.00  14410.50  14412.00   44.0\n",
       "2023-03-21 15:08:45+00:00  12742.00  12742.75  12739.50  12740.75   45.0\n",
       "2023-04-28 06:04:30+00:00  13216.25  13216.25  13215.25  13215.50   46.0\n",
       "2023-05-11 18:22:15+00:00  13424.00  13428.00  13424.00  13427.75   50.0\n",
       "2023-05-29 04:59:45+00:00  14394.25  14395.25  14394.00  14395.00   51.0\n",
       "2023-03-28 05:27:00+00:00  12796.25  12796.75  12796.00  12796.50   52.0\n",
       "2023-06-01 18:28:00+00:00  14493.75  14494.75  14493.00  14494.00   53.0\n",
       "2023-03-13 19:22:15+00:00  12124.50  12126.00  12120.25  12123.25   54.0\n",
       "2023-04-03 14:43:30+00:00  13238.75  13239.25  13236.00  13237.25   55.0\n",
       "2023-05-04 01:44:45+00:00  13128.50  13128.50  13127.75  13128.00   56.0\n",
       "2023-06-13 00:48:45+00:00  14816.50  14817.00  14816.50  14817.00   60.0\n",
       "2023-06-06 01:01:30+00:00  14573.75  14575.00  14573.25  14575.00   61.0\n",
       "2023-05-03 17:25:15+00:00  13223.75  13225.00  13223.25  13224.75   62.0\n",
       "2023-06-09 11:57:15+00:00  14516.50  14517.00  14515.75  14516.75   63.0\n",
       "2023-03-23 07:25:00+00:00  12812.25  12812.50  12811.50  12812.25   64.0\n",
       "2023-06-08 18:08:15+00:00  14443.50  14443.50  14441.25  14442.75   65.0\n",
       "2023-05-08 19:29:00+00:00  13345.25  13345.25  13340.25  13343.25   66.0\n",
       "2023-03-29 04:24:15+00:00  12778.00  12778.25  12777.75  12778.25   72.0\n",
       "2023-04-12 05:46:30+00:00  13076.25  13076.50  13075.75  13076.50   73.0\n",
       "2023-05-26 06:06:15+00:00  13963.00  13963.25  13962.25  13963.25   74.0\n",
       "2023-04-10 11:00:45+00:00  13152.00  13152.00  13151.00  13151.75   75.0\n",
       "2023-05-29 07:14:00+00:00  14409.50  14409.75  14408.25  14409.75   84.0\n",
       "2023-05-28 23:42:45+00:00  14419.00  14419.00  14418.75  14419.00   85.0"
      ]
     },
     "execution_count": 508,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels = torch.round(model0.label.cpu().detach(),decimals=0)\n",
    "uniq = torch.unique(labels,dim=0)\n",
    "diag = torch.tensor([10.0,1])\n",
    "code_list = {}\n",
    "for i,code in enumerate((uniq @ diag.view(2,1)).numpy()):\n",
    "    code_list[code.item()]= i\n",
    "codes = (labels @ diag.view(2,1))\n",
    "df1['codes'] = codes\n",
    "df1.groupby('codes').sample()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 509,
   "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>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>codes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2023-03-13 04:00:00+00:00</th>\n",
       "      <td>12177.25</td>\n",
       "      <td>12179.00</td>\n",
       "      <td>12177.25</td>\n",
       "      <td>12179.00</td>\n",
       "      <td>29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-13 04:00:15+00:00</th>\n",
       "      <td>12179.00</td>\n",
       "      <td>12179.25</td>\n",
       "      <td>12179.00</td>\n",
       "      <td>12179.25</td>\n",
       "      <td>29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-13 04:00:30+00:00</th>\n",
       "      <td>12181.25</td>\n",
       "      <td>12181.25</td>\n",
       "      <td>12180.25</td>\n",
       "      <td>12180.75</td>\n",
       "      <td>28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-13 04:00:45+00:00</th>\n",
       "      <td>12179.50</td>\n",
       "      <td>12179.75</td>\n",
       "      <td>12179.25</td>\n",
       "      <td>12179.75</td>\n",
       "      <td>36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-03-13 04:01:00+00:00</th>\n",
       "      <td>12180.00</td>\n",
       "      <td>12181.75</td>\n",
       "      <td>12180.00</td>\n",
       "      <td>12180.50</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-13 03:58:45+00:00</th>\n",
       "      <td>14865.00</td>\n",
       "      <td>14865.00</td>\n",
       "      <td>14864.50</td>\n",
       "      <td>14864.50</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-13 03:59:00+00:00</th>\n",
       "      <td>14864.25</td>\n",
       "      <td>14865.00</td>\n",
       "      <td>14864.25</td>\n",
       "      <td>14865.00</td>\n",
       "      <td>29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-13 03:59:15+00:00</th>\n",
       "      <td>14864.50</td>\n",
       "      <td>14866.00</td>\n",
       "      <td>14864.50</td>\n",
       "      <td>14866.00</td>\n",
       "      <td>29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-13 03:59:30+00:00</th>\n",
       "      <td>14866.75</td>\n",
       "      <td>14867.00</td>\n",
       "      <td>14866.50</td>\n",
       "      <td>14866.50</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-13 03:59:45+00:00</th>\n",
       "      <td>14867.25</td>\n",
       "      <td>14867.25</td>\n",
       "      <td>14865.00</td>\n",
       "      <td>14866.25</td>\n",
       "      <td>35</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>354167 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                               open      high       low     close  codes\n",
       "2023-03-13 04:00:00+00:00  12177.25  12179.00  12177.25  12179.00     29\n",
       "2023-03-13 04:00:15+00:00  12179.00  12179.25  12179.00  12179.25     29\n",
       "2023-03-13 04:00:30+00:00  12181.25  12181.25  12180.25  12180.75     28\n",
       "2023-03-13 04:00:45+00:00  12179.50  12179.75  12179.25  12179.75     36\n",
       "2023-03-13 04:01:00+00:00  12180.00  12181.75  12180.00  12180.50      3\n",
       "...                             ...       ...       ...       ...    ...\n",
       "2023-06-13 03:58:45+00:00  14865.00  14865.00  14864.50  14864.50     14\n",
       "2023-06-13 03:59:00+00:00  14864.25  14865.00  14864.25  14865.00     29\n",
       "2023-06-13 03:59:15+00:00  14864.50  14866.00  14864.50  14866.00     29\n",
       "2023-06-13 03:59:30+00:00  14866.75  14867.00  14866.50  14866.50      6\n",
       "2023-06-13 03:59:45+00:00  14867.25  14867.25  14865.00  14866.25     35\n",
       "\n",
       "[354167 rows x 5 columns]"
      ]
     },
     "execution_count": 509,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1['codes'] = df1['codes'].apply(lambda x : code_list[x])\n",
    "df1\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 573,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "plotlyServerURL": "https://plot.ly"
       },
       "data": [
        {
         "close": [
          0,
          0.1111111111111111,
          0,
          0.2857142857142857,
          0.14285714285714285,
          0,
          0,
          0,
          0.5,
          0.3333333333333333,
          0.21052631578947367,
          0.3125,
          0.125,
          0,
          0,
          0.75,
          0.6923076923076923,
          0.5833333333333334,
          0.5,
          0.3333333333333333,
          0.18181818181818182,
          0.125,
          0.7777777777777778,
          0.8928571428571429,
          0.6666666666666666,
          0.6666666666666666,
          0.5384615384615384,
          0.5,
          0.38461538461538464,
          1,
          1,
          0.8823529411764706,
          0.75,
          0.6666666666666666,
          0.6,
          0.6,
          1,
          1,
          0.8333333333333334,
          0.75,
          1,
          1
         ],
         "high": [
          0,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1,
          1
         ],
         "low": [
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0,
          0
         ],
         "open": [
          0,
          0,
          0.2,
          0,
          0.14285714285714285,
          0.3333333333333333,
          0.4,
          0.5714285714285714,
          0,
          0.1111111111111111,
          0.42105263157894735,
          0.5,
          0.6875,
          0.8,
          1,
          0,
          0,
          0.16666666666666666,
          0.4,
          0.6666666666666666,
          0.8181818181818182,
          1,
          0,
          0.10714285714285714,
          0.3333333333333333,
          0.4444444444444444,
          0.6153846153846154,
          0.75,
          1,
          0,
          0.25,
          0.4117647058823529,
          0.5,
          0.6666666666666666,
          0.8,
          1,
          0.4,
          0.6666666666666666,
          0.8333333333333334,
          1,
          0.8571428571428571,
          1
         ],
         "type": "candlestick",
         "x": [
          0,
          1,
          2,
          3,
          4,
          5,
          6,
          7,
          8,
          9,
          10,
          11,
          12,
          13,
          14,
          15,
          16,
          17,
          18,
          19,
          20,
          21,
          22,
          23,
          24,
          25,
          26,
          27,
          28,
          29,
          30,
          31,
          32,
          33,
          34,
          35,
          36,
          37,
          38,
          39,
          40,
          41
         ]
        }
       ],
       "layout": {
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "fillpattern": {
             "fillmode": "overlay",
             "size": 10,
             "solidity": 0.2
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "autotypenumbers": "strict",
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        },
        "xaxis": {
         "rangeslider": {
          "visible": false
         }
        }
       }
      }
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_sample(MinMaxNorm((torch.tensor(df1.groupby('codes').sample().iloc[:,:4].values)),1))\n"
   ]
  }
 ],
 "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": "3.11.4"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}