HAR-combined_final.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import os\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn import preprocessing\n",
    "from sklearn.model_selection import cross_val_score\n",
    "import math\n",
    "from scipy.stats import iqr\n",
    "from numpy import linalg\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import accuracy_score\n",
    "from scipy import stats\n",
    "import time\n",
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## Features\n",
    "\n",
    "def mean(x, y, z):\n",
    "    \"\"\"Calculates mean\"\"\"\n",
    "    mean_x = np.mean(x)\n",
    "    mean_y = np.mean(y)\n",
    "    mean_z = np.mean(z)\n",
    "    return mean_x, mean_y, mean_z\n",
    "\n",
    "def std_dev(x, y, z):\n",
    "    \"\"\"Calculates standard deviation\"\"\"\n",
    "    std_x = np.std(x)\n",
    "    std_y = np.std(y)\n",
    "    std_z = np.std(z)\n",
    "    return std_x, std_y, std_z  \n",
    "    \n",
    "def mad(x, y, z):\n",
    "    \"\"\"Calculates median absolute deviation\"\"\"\n",
    "    mad_x = np.median(np.abs(x - np.median(x)))\n",
    "    mad_y = np.median(np.abs(y - np.median(y)))\n",
    "    mad_z = np.median(np.abs(z - np.median(z)))\n",
    "    return mad_x, mad_y, mad_z\n",
    "\n",
    "def minimum(x, y, z):\n",
    "    \"\"\"Calculates minimum\"\"\"\n",
    "    return min(x), min(y), min(z)\n",
    "\n",
    "def maximum(x, y, z):\n",
    "    \"\"\"Calculates maximum\"\"\"\n",
    "    return max(x), max(y), max(z)\n",
    "\n",
    "def energy_measure(x, y, z):\n",
    "    \"\"\"Calculates energy measures\"\"\"\n",
    "    em_x = np.mean(np.square(x))\n",
    "    em_y = np.mean(np.square(y))\n",
    "    em_z = np.mean(np.square(z))\n",
    "    return em_x, em_y, em_z\n",
    "\n",
    "def inter_quartile_range(x, y, z):\n",
    "    \"\"\"Calculates inter-quartile range\"\"\"\n",
    "    iqr_x = np.subtract(*np.percentile(x, [75, 25]))\n",
    "    iqr_y = np.subtract(*np.percentile(y, [75, 25]))\n",
    "    iqr_z = np.subtract(*np.percentile(z, [75, 25]))\n",
    "    return iqr_x, iqr_y, iqr_z\n",
    "\n",
    "def sma(x, y, z):\n",
    "    \"\"\"Calculates signal magnitude area\"\"\"   \n",
    "    abs_x = np.absolute(x)\n",
    "    abs_y = np.absolute(y)\n",
    "    abs_z = np.absolute(z)\n",
    "    return np.mean(abs_x+abs_y+abs_z)\n",
    "\n",
    "def skewness(x, y, z):\n",
    "    \"\"\"Calculates skewness\"\"\"\n",
    "    skew_x=stats.skew(x)\n",
    "    skew_y=stats.skew(y)\n",
    "    skew_z=stats.skew(z)\n",
    "    return skew_x, skew_y, skew_z\n",
    "\n",
    "def kurt(x, y, z):\n",
    "    \"\"\"Calculates kurtosis\"\"\"\n",
    "    kurt_x = stats.kurtosis(x,fisher=False)\n",
    "    kurt_y = stats.kurtosis(y,fisher=False)\n",
    "    kurt_z = stats.kurtosis(z,fisher=False)\n",
    "    return kurt_x, kurt_y, kurt_z\n",
    "    \n",
    "def max_index(x, y, z):\n",
    "    \"\"\"Calculates index of maximum\"\"\"\n",
    "    idx_x = np.argmax(x)\n",
    "    idx_y = np.argmax(y)\n",
    "    idx_z = np.argmax(z)\n",
    "    return idx_x, idx_y, idx_z\n",
    "\n",
    "def butter_bandpass(lowcut, highcut, fs, order=5):\n",
    "    nyq = 0.5 * fs\n",
    "    low = lowcut / nyq\n",
    "    high = highcut / nyq\n",
    "    b, a = butter(order, [low, high], btype='band')\n",
    "    return b, a\n",
    "\n",
    "\n",
    "def butter_bandpass_filter(data, lowcut, highcut, fs, order=5):\n",
    "    b, a = butter_bandpass(lowcut, highcut, fs, order=order)\n",
    "    y = lfilter(b, a, data)\n",
    "    return y\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def process_acc(input_df, columns, output_df, freq_flag):\n",
    "    i = 0\n",
    "    while i < len(input_df[0]):\n",
    "        max_window = input_df[0][i] + 10000\n",
    "        min_window = input_df[0][i]\n",
    "        if i != 0:                                  #handling overlap\n",
    "            min_window -= 5000\n",
    "            max_window -= 5000\n",
    "            i -= int(count/2)\n",
    "        x = [input_df[1][i]]\n",
    "        y = [input_df[2][i]]\n",
    "        z = [input_df[3][i]]\n",
    "        \n",
    "        count = 0\n",
    "        while min_window < max_window:\n",
    "            i += 1\n",
    "            count += 1\n",
    "            try:\n",
    "                label = input_df[5][i]\n",
    "                if label=='null' or label is None:\n",
    "                    continue\n",
    "                x.append(input_df[1][i])\n",
    "                y.append(input_df[2][i])\n",
    "                z.append(input_df[3][i])\n",
    "                min_window = input_df[0][i]\n",
    "            except KeyError:\n",
    "                break\n",
    "\n",
    "        sma_val = (sma(x,y,z), )\n",
    "        \n",
    "        if freq_flag == 0:\n",
    "            acc_data = list(mean(x, y, z)+std_dev(x, y, z)+mad(x, y, z)+minimum(x, y, z)+maximum(x, y, z)+energy_measure(x, y, z)+inter_quartile_range(x, y, z)+sma_val)\n",
    "        \n",
    "        \n",
    "        else:\n",
    "            acc_data = list(mean(x, y, z)+std_dev(x, y, z)+mad(x, y, z)+minimum(x, y, z)+maximum(x, y, z)+energy_measure(x, y, z)+inter_quartile_range(x, y, z)+sma_val+\n",
    "                            skewness(x, y, z)+kurt(x, y, z)+max_index(x, y, z))\n",
    "        \n",
    "        \n",
    "       # acc_data = list(mean(x, y, z)+std_dev(x, y, z)+mad(x, y, z)+minimum(x, y, z)+maximum(x, y, z)+energy_measure(x, y, z)+inter_quartile_range(x, y, z)+sma_val)\n",
    "        output_df = output_df.append(pd.Series(acc_data, index=columns), ignore_index=True)\n",
    "\n",
    "    return output_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def process_gyr(input_df, columns, output_df, freq_flag):\n",
    "\n",
    "    i = 0\n",
    "    while i < len(input_df[0]):\n",
    "        max_window = input_df[0][i] + 10000\n",
    "        min_window = input_df[0][i]\n",
    "        if i != 0:                                                                 #handling overlap\n",
    "            min_window -= 5000\n",
    "            max_window -= 5000\n",
    "            i -= int(count/2)\n",
    "        x = [input_df[1][i]]\n",
    "        y = [input_df[2][i]]\n",
    "        z = [input_df[3][i]]\n",
    "        l = input_df[5][i]\n",
    "        \n",
    "        count = 0\n",
    "        while min_window < max_window:\n",
    "            i += 1\n",
    "            count += 1\n",
    "            try:\n",
    "                label = input_df[5][i]\n",
    "                if label=='null' or label is None:\n",
    "                    continue\n",
    "                x.append(input_df[1][i])\n",
    "                y.append(input_df[2][i])\n",
    "                z.append(input_df[3][i])\n",
    "                min_window = input_df[0][i]\n",
    "            except KeyError:\n",
    "                break\n",
    "        \n",
    "        l = (l, )  \n",
    "        sma_val = (sma(x,y,z), )\n",
    "        \n",
    "        if freq_flag == 0:\n",
    "            gy_data = list(mean(x, y, z)+std_dev(x, y, z)+mad(x, y, z)+minimum(x, y, z)+maximum(x, y, z)+energy_measure(x, y, z)+inter_quartile_range(x, y, z)+sma_val+l)\n",
    "        \n",
    "        \n",
    "        else:\n",
    "            gy_data = list(mean(x, y, z)+std_dev(x, y, z)+mad(x, y, z)+minimum(x, y, z)+maximum(x, y, z)+energy_measure(x, y, z)+inter_quartile_range(x, y, z)+sma_val+\n",
    "                            skewness(x, y, z)+kurt(x, y, z)+max_index(x, y, z)+l)\n",
    "        \n",
    "        #gy_data = list(mean(x, y, z)+std_dev(x, y, z)+mad(x, y, z)+minimum(x, y, z)+maximum(x, y, z)+energy_measure(x, y, z)+inter_quartile_range(x, y, z)+sma_val+label) \n",
    "        output_df = output_df.append(pd.Series(gy_data, index=columns), ignore_index=True)\n",
    "        \n",
    "    return output_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def process_mag(input_df, columns, output_df, freq_flag):\n",
    "    mag = np.linalg.norm(input_df.iloc[:,1:4], axis = 1)\n",
    "    input_df[6] = mag.tolist()\n",
    "    \n",
    "    i = 0\n",
    "    while i < len(input_df[0]):\n",
    "        max_window = input_df[0][i] + 10000\n",
    "        min_window = input_df[0][i]\n",
    "        if i != 0:                                  #handling overlap\n",
    "            min_window -= 5000\n",
    "            max_window -= 5000\n",
    "            i -= int(count/2)\n",
    "        x = [input_df[6][i]]\n",
    "        \n",
    "        \n",
    "        count = 0\n",
    "        while min_window < max_window:\n",
    "            i += 1\n",
    "            count += 1\n",
    "            try:\n",
    "                label = input_df[5][i]\n",
    "                if label=='null' or label is None:\n",
    "                    continue\n",
    "                x.append(input_df[6][i])\n",
    "                min_window = input_df[0][i]\n",
    "            except KeyError:\n",
    "                break\n",
    "                \n",
    "        label = (input_df[5][1], )\n",
    "        if label == 'null':\n",
    "            print('--------null')\n",
    "            return\n",
    "        \n",
    "        mean = (np.mean(x),)\n",
    "        std_dev = (np.std(x),)\n",
    "        mad = (np.median(np.abs(x - np.median(x))),)\n",
    "        minimum = (np.min(x),)\n",
    "        maximum = (np.max(x),)\n",
    "        energy_measure = (np.mean(np.square(x)),)\n",
    "        inter_quartile_range = (np.subtract(*np.percentile(x, [75, 25])),)\n",
    "        sma_val = (np.mean(np.absolute(x)),)\n",
    "        \n",
    "        if freq_flag == 0:\n",
    "            acc_data = list(mean+std_dev+mad+minimum+maximum+energy_measure+inter_quartile_range+sma_val)\n",
    "        else:\n",
    "            skew = (stats.skew(x),)\n",
    "            kur = (stats.kurtosis(x,fisher=False),)\n",
    "            m_idx = (np.argmax(x),)\n",
    "            acc_data = list(mean+std_dev+mad+minimum+maximum+energy_measure+inter_quartile_range+sma_val+skew+kur+m_idx)\n",
    "        \n",
    "        #acc_data = list(mean+std_dev+mad+minimum+maximum+energy_measure+inter_quartile_range+sma_val)\n",
    "        output_df = output_df.append(pd.Series(acc_data, index=columns), ignore_index=True)\n",
    "\n",
    "    return output_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def hist_activities(df):\n",
    "    df['label'].value_counts().plot(kind='bar', title='Training examples by activity type')\n",
    "    plt.show()\n",
    "    \n",
    "def plot_activity(activity, df):\n",
    "    data = df[df.iloc[:, 5] == activity][[1, 2, 3]][:2000]\n",
    "    axis = data.plot(subplots=True, figsize=(16, 12), \n",
    "                     title=activity)\n",
    "    for ax in axis:\n",
    "        ax.legend(loc='lower left', bbox_to_anchor=(1.0, 0.5))\n",
    "        \n",
    "def plot_act_signals(df):\n",
    "    plot_activity(\"sitting\", df)\n",
    "    plt.ylabel('sitting')\n",
    "    plt.show()\n",
    "    plot_activity(\"walking\", df)\n",
    "    plt.ylabel('walking')\n",
    "    plt.show()\n",
    "    plot_activity(\"standing\", df)\n",
    "    plt.ylabel('standing')\n",
    "    plt.show()\n",
    "    plot_activity(\"laying_down\", df)\n",
    "    plt.ylabel('laying_down')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def calc_jerk(acc, ts):\n",
    "    jk = [0]* len(acc)\n",
    "    for i in range(1,len(acc)):\n",
    "        jk[i-1] = 1000*(acc[i] - acc[i-1])/(ts[i] - ts[i-1])\n",
    "    return jk\n",
    "\n",
    "def fmagnitude(num):\n",
    "    return math.sqrt(num.real**2 + num.imag**2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_session_dirs(path):\n",
    "    path = path\n",
    "    dirs = []\n",
    "    dir_list = [x[0] for x in os.walk(path)]\n",
    "    last_dirs = [\"1_android.sensor.accelerometer.data.csv.gz\",\"4_android.sensor.gyroscope.data.csv.gz\"]\n",
    "    dirs = []\n",
    "    for each_dir in dir_list[2:]:\n",
    "\n",
    "        if \"data\" in each_dir:\n",
    "            dirs.append(each_dir)\n",
    "\n",
    "    dirs.sort(key=lambda x:x[-22:])\n",
    "    return dirs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_jerk(df):\n",
    "    df[1] = calc_jerk(df[1],df[0])\n",
    "    df[2] = calc_jerk(df[2],df[0])\n",
    "    df[3] = calc_jerk(df[3],df[0])\n",
    "    return df\n",
    "\n",
    "def get_fft(df):\n",
    "    df[1] = np.fft.fft(df[1])\n",
    "    df[2] = np.fft.fft(df[2])\n",
    "    df[3] = np.fft.fft(df[3])\n",
    "    df[1] = [fmagnitude(j) for j in df[1]]\n",
    "    df[2] = [fmagnitude(j) for j in df[2]]\n",
    "    df[3] = [fmagnitude(j) for j in df[3]]\n",
    "    return df\n",
    "\n",
    "def five_point_smoothing(df):\n",
    "    m=np.array(df[1])\n",
    "    m_smooth = np.zeros(len(m))\n",
    "    m_smooth = m;\n",
    "    for i ,val in enumerate(m_smooth[2 : (len(m_smooth) -2)]):\n",
    "        m_smooth[i] = (m_smooth[i-2] + m_smooth[i-1] + m_smooth[i] + m_smooth[i+1] + m_smooth[i+2])/5\n",
    "    df[1] = m_smooth\n",
    "        \n",
    "    m=np.array(df[2])\n",
    "    m_smooth = np.zeros(len(m))\n",
    "    m_smooth = m;\n",
    "    for i ,val in enumerate(m_smooth[2 : (len(m_smooth) -2)]):\n",
    "        m_smooth[i] = (m_smooth[i-2] + m_smooth[i-1] + m_smooth[i] + m_smooth[i+1] + m_smooth[i+2])/5\n",
    "    df[2] = m_smooth\n",
    "    \n",
    "    m = np.array(df[3])\n",
    "    m_smooth = np.zeros(len(m))\n",
    "    m_smooth = m;\n",
    "    for i ,val in enumerate(m_smooth[2 : (len(m_smooth) -2)]):\n",
    "        m_smooth[i] = (m_smooth[i-2] + m_smooth[i-1] + m_smooth[i] + m_smooth[i+1] + m_smooth[i+2])/5\n",
    "    df[3] = m_smooth\n",
    "    \n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "tacc_cols = ['mean_acc_x', 'mean_acc_y', 'mean_acc_z', \n",
    "               'std_acc_x', 'std_acc_y', 'std_acc_z', \n",
    "               'mad_acc_x', 'mad_acc_y', 'mad_acc_z',\n",
    "               'min_acc_x', 'min_acc_y', 'min_acc_z', \n",
    "               'max_acc_x', 'max_acc_y', 'max_acc_z', \n",
    "               'em_acc_x', 'em_acc_y', 'em_acc_z', \n",
    "               'iqr_acc_x', 'iqr_acc_y', 'iqr_acc_z', \n",
    "               'sma_acc']\n",
    "\n",
    "tacc_jerk_cols = ['mean_accj_x', 'mean_accj_y', 'mean_accj_z', \n",
    "                   'std_accj_x', 'std_accj_y', 'std_accj_z', \n",
    "                   'mad_accj_x', 'mad_accj_y', 'mad_accj_z',\n",
    "                   'min_accj_x', 'min_accj_y', 'min_accj_z', \n",
    "                   'max_accj_x', 'max_accj_y', 'max_accj_z', \n",
    "                   'em_accj_x', 'em_accj_y', 'em_accj_z', \n",
    "                   'iqr_accj_x', 'iqr_accj_y', 'iqr_accj_z', \n",
    "                   'sma_accj']\n",
    "\n",
    "tacc_mag_cols = ['mean_macc', 'std_macc', 'mad_macc', 'min_macc', 'max_macc', 'em_macc', 'iqr_macc', 'sma_macc']\n",
    "\n",
    "tacc_jmag_cols = ['mean_macc_jerk', 'std_macc_jerk', 'mad_macc_jerk', 'min_macc_jerk', 'max_macc_jerk', \n",
    "                     'em_macc_jerk', 'iqr_macc_jerk', 'sma_macc_jerk']\n",
    "\n",
    "facc_cols = ['mean_facc_x', 'mean_facc_y', 'mean_facc_z', \n",
    "               'std_facc_x', 'std_facc_y', 'std_facc_z', \n",
    "               'mad_facc_x', 'mad_facc_y', 'mad_facc_z',\n",
    "               'min_facc_x', 'min_facc_y', 'min_facc_z', \n",
    "               'max_facc_x', 'max_facc_y', 'max_facc_z', \n",
    "               'em_facc_x', 'em_facc_y', 'em_facc_z', \n",
    "               'iqr_facc_x', 'iqr_facc_y', 'iqr_facc_z', \n",
    "               'sma_facc',\n",
    "               'skew_facc_x', 'skew_facc_y', 'skew_facc_z', \n",
    "               'kurt_facc_x', 'kurt_facc_y', 'kurt_facc_z', \n",
    "               'idx_facc_x', 'idx_facc_y', 'idx_facc_z']\n",
    "\n",
    "facc_jerk_cols = ['mean_faccj_x', 'mean_faccj_y', 'mean_faccj_z', \n",
    "                   'std_faccj_x', 'std_faccj_y', 'std_faccj_z', \n",
    "                   'mad_faccj_x', 'mad_faccj_y', 'mad_faccj_z',\n",
    "                   'min_faccj_x', 'min_faccj_y', 'min_faccj_z', \n",
    "                   'max_faccj_x', 'max_faccj_y', 'max_faccj_z', \n",
    "                   'em_faccj_x', 'em_faccj_y', 'em_faccj_z', \n",
    "                   'iqr_faccj_x', 'iqr_faccj_y', 'iqr_faccj_z', \n",
    "                   'sma_faccj',\n",
    "                   'skew_faccj_x', 'skew_faccj_y', 'skew_faccj_z', \n",
    "                   'kurt_faccj_x', 'kurt_faccj_y', 'kurt_faccj_z', \n",
    "                   'idx_faccj_x', 'idx_faccj_y', 'idx_faccj_z']\n",
    "\n",
    "facc_mag_cols = ['mean_mfacc', 'std_mfacc', 'mad_mfacc', 'min_mfacc', 'max_mfacc', 'em_mfacc', 'iqr_mfacc', 'sma_mfacc',\n",
    "                'skew_mfacc', 'kurt_mfacc',  'idx_mfacc']\n",
    "\n",
    "facc_jmag_cols = ['mean_mfacc_jerk', 'std_mfacc_jerk', 'mad_mfacc_jerk', 'min_mfacc_jerk', 'max_mfacc_jerk', \n",
    "                     'em_mfacc_jerk', 'iqr_mfacc_jerk', 'sma_mfacc_jerk', 'skew_mfacc_jerk', 'kurt_mfacc_jerk',  \n",
    "                  'idx_mfacc_jerk']\n",
    "\n",
    "tgy_cols = ['mean_gy_x', 'mean_gy_y', 'mean_gy_z', \n",
    "               'std_gy_x', 'std_gy_y', 'std_gy_z', \n",
    "               'mad_gy_x', 'mad_gy_y', 'mad_gy_z',\n",
    "               'min_gy_x', 'min_gy_y', 'min_gy_z', \n",
    "               'max_gy_x', 'max_gy_y', 'max_gy_z', \n",
    "               'em_gy_x', 'em_gy_y', 'em_gy_z', \n",
    "               'iqr_gy_x', 'iqr_gy_y', 'iqr_gy_z', \n",
    "               'sma_gy', 'label']\n",
    "\n",
    "tgy_jerk_cols = ['mean_gyj_x', 'mean_gyj_y', 'mean_gyj_z', \n",
    "                   'std_gyj_x', 'std_gyj_y', 'std_gyj_z', \n",
    "                   'mad_gyj_x', 'mad_gyj_y', 'mad_gyj_z',\n",
    "                   'min_gyj_x', 'min_gyj_y', 'min_gyj_z', \n",
    "                   'max_gyj_x', 'max_gyj_y', 'max_gyj_z', \n",
    "                   'em_gyj_x', 'em_gyj_y', 'em_gyj_z', \n",
    "                   'iqr_gyj_x', 'iqr_gyj_y', 'iqr_gyj_z', \n",
    "                   'sma_gyj']\n",
    "\n",
    "tgy_mag_cols = ['mean_mgy', 'std_mgy', 'mad_mgy', 'min_mgy', 'max_mgy', 'em_mgy', 'iqr_mgy', 'sma_mgy']\n",
    "\n",
    "tgy_jmag_cols = ['mean_mgy_jerk', 'std_mgy_jerk', 'mad_mgy_jerk', 'min_mgy_jerk', 'max_mgy_jerk', \n",
    "                     'em_mgy_jerk', 'iqr_mgy_jerk', 'sma_mgy_jerk']\n",
    "fgy_cols = ['mean_fgy_x', 'mean_fgy_y', 'mean_fgy_z', \n",
    "               'std_fgy_x', 'std_fgy_y', 'std_fgy_z', \n",
    "               'mad_fgy_x', 'mad_fgy_y', 'mad_fgy_z',\n",
    "               'min_fgy_x', 'min_fgy_y', 'min_fgy_z', \n",
    "               'max_fgy_x', 'max_fgy_y', 'max_fgy_z', \n",
    "               'em_fgy_x', 'em_fgy_y', 'em_fgy_z', \n",
    "               'iqr_fgy_x', 'iqr_fgy_y', 'iqr_fgy_z', \n",
    "               'sma_fgy', \n",
    "               'skew_fgy_x', 'skew_fgy_y', 'skew_fgy_z', \n",
    "               'kurt_fgy_x', 'kurt_fgy_y', 'kurt_fgy_z', \n",
    "               'idx_fgy_x', 'idx_fgy_y', 'idx_fgy_z']\n",
    "\n",
    "fgy_jerk_cols = ['mean_fgyj_x', 'mean_fgyj_y', 'mean_fgyj_z', \n",
    "                   'std_fgyj_x', 'std_fgyj_y', 'std_fgyj_z', \n",
    "                   'mad_fgyj_x', 'mad_fgyj_y', 'mad_fgyj_z',\n",
    "                   'min_fgyj_x', 'min_fgyj_y', 'min_fgyj_z', \n",
    "                   'max_fgyj_x', 'max_fgyj_y', 'max_fgyj_z', \n",
    "                   'em_fgyj_x', 'em_fgyj_y', 'em_fgyj_z', \n",
    "                   'iqr_fgyj_x', 'iqr_fgyj_y', 'iqr_fgyj_z', \n",
    "                   'sma_fgyj',\n",
    "                   'skew_fgyj_x', 'skew_fgyj_y', 'skew_fgyj_z', \n",
    "                   'kurt_fgyj_x', 'kurt_fgyj_y', 'kurt_fgyj_z', \n",
    "                   'idx_fgyj_x', 'idx_fgyj_y', 'idx_fgyj_z']\n",
    "\n",
    "fgy_mag_cols = ['mean_mfgy', 'std_mfgy', 'mad_mfgy', 'min_mfgy', 'max_mfgy', 'em_mfgy', 'iqr_mfgy', 'sma_mfgy',\n",
    "               'skew_mfgy', 'kurt_mfgy',  'idx_mfgy']\n",
    "\n",
    "fgy_jmag_cols = ['mean_mfgy_jerk', 'std_mfgy_jerk', 'mad_mfgy_jerk', 'min_mfgy_jerk', 'max_mfgy_jerk', \n",
    "                     'em_mfgy_jerk', 'iqr_mfgy_jerk', 'sma_mfgy_jerk', 'skew_mfgy_jerk', 'kurt_mfgy_jerk',  'idx_mfgy_jerk']\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def process_csv_data(path):\n",
    "    dirs = get_session_dirs(path)\n",
    "\n",
    "    out_acc_df = pd.DataFrame(columns = tacc_cols)\n",
    "\n",
    "    out_acc_jerk_df = pd.DataFrame(columns = tacc_jerk_cols)\n",
    "\n",
    "    acc_mag = pd.DataFrame(columns = tacc_mag_cols)\n",
    "\n",
    "    acc_jerk_mag = pd.DataFrame(columns = tacc_jmag_cols)\n",
    "\n",
    "    out_facc_df = pd.DataFrame(columns = facc_cols)\n",
    "\n",
    "    out_facc_jerk_df = pd.DataFrame(columns = facc_jerk_cols)\n",
    "\n",
    "    facc_mag = pd.DataFrame(columns = facc_mag_cols)\n",
    "\n",
    "    facc_jerk_mag = pd.DataFrame(columns = facc_jmag_cols)\n",
    "\n",
    "    out_gy_df = pd.DataFrame(columns = tgy_cols)\n",
    "\n",
    "    out_gy_jerk_df = pd.DataFrame(columns = tgy_jerk_cols)\n",
    "\n",
    "    gy_mag = pd.DataFrame(columns = tgy_mag_cols)\n",
    "\n",
    "    gy_jerk_mag = pd.DataFrame(columns = tgy_jmag_cols)\n",
    "\n",
    "    out_fgy_df = pd.DataFrame(columns = fgy_cols)\n",
    "\n",
    "    out_fgy_jerk_df = pd.DataFrame(columns = fgy_jerk_cols)\n",
    "\n",
    "    fgy_mag = pd.DataFrame(columns = fgy_mag_cols)\n",
    "\n",
    "    fgy_jerk_mag = pd.DataFrame(columns = fgy_jmag_cols)\n",
    "\n",
    "    acc_gy_df = pd.DataFrame()\n",
    "\n",
    "    for i in tqdm(range(len(dirs))): \n",
    "        full_path_acc = os.path.normpath(dirs[i] + \"/\" + \"1_android.sensor.accelerometer.data.csv.gz\")\n",
    "        full_path_gy = os.path.normpath(dirs[i] + \"/\" + \"4_android.sensor.gyroscope.data.csv.gz\")\n",
    "    \n",
    "\n",
    "        print(\"processing acceleration: \" + full_path_acc)\n",
    "        in_df_ac = pd.read_csv(full_path_acc, compression='gzip', sep=',', header=None)\n",
    "        \n",
    "        if in_df_ac.shape[0] > 3000:\n",
    "    \n",
    "            in_df_ac = in_df_ac.drop(list(range(1000)), axis = 0)\n",
    "            in_df_ac.index = range(in_df_ac.shape[0])\n",
    "            in_df_ac = in_df_ac.drop(list(range(in_df_ac.shape[0]-1000, in_df_ac.shape[0])), axis = 0)\n",
    "            in_df_ac.index = range(in_df_ac.shape[0])\n",
    "\n",
    "            \n",
    "        in_df_ac = five_point_smoothing(in_df_ac)\n",
    "    \n",
    "        in_df_ac_jerk = in_df_ac.copy()\n",
    "        in_df_ac_jerk = get_jerk(in_df_ac_jerk)\n",
    "\n",
    "        in_df_fac = in_df_ac.copy()\n",
    "        in_df_fac = get_fft(in_df_fac)\n",
    "\n",
    "        in_df_fac_jerk = in_df_ac_jerk.copy()\n",
    "        in_df_fac_jerk = get_fft(in_df_fac_jerk)\n",
    "\n",
    "        out_acc_df = process_acc(in_df_ac, tacc_cols, out_acc_df, 0)\n",
    "        out_acc_jerk_df = process_acc(in_df_ac_jerk, tacc_jerk_cols, out_acc_jerk_df, 0)\n",
    "        out_facc_df = process_acc(in_df_fac, facc_cols, out_facc_df, 1)\n",
    "        out_facc_jerk_df = process_acc(in_df_fac_jerk, facc_jerk_cols, out_facc_jerk_df, 1)\n",
    "\n",
    "        acc_mag = process_mag(in_df_ac, tacc_mag_cols, acc_mag, 0)\n",
    "        acc_jerk_mag = process_mag(in_df_ac_jerk, tacc_jmag_cols, acc_jerk_mag, 0)\n",
    "        facc_mag = process_mag(in_df_fac, facc_mag_cols, facc_mag, 1)\n",
    "        facc_jerk_mag = process_mag(in_df_fac_jerk, facc_jmag_cols, facc_jerk_mag, 1)\n",
    "\n",
    "        print(\"processing gyroscope: \" + full_path_gy)\n",
    "        in_df_gy = pd.read_csv(full_path_gy, compression='gzip', sep=',', header=None)\n",
    "        \n",
    "        if in_df_gy.shape[0] > 3000:\n",
    "        \n",
    "            in_df_gy = in_df_gy.drop(list(range(1000)), axis = 0)\n",
    "            in_df_gy.index = range(in_df_gy.shape[0])\n",
    "            in_df_gy = in_df_gy.drop(list(range(in_df_gy.shape[0]-1000, in_df_gy.shape[0])), axis = 0)\n",
    "            in_df_gy.index = range(in_df_gy.shape[0])\n",
    "\n",
    "        in_df_gy = five_point_smoothing(in_df_gy)\n",
    "        \n",
    "        in_df_gy_jerk = in_df_gy.copy()\n",
    "        in_df_gy_jerk = get_jerk(in_df_gy_jerk)\n",
    "\n",
    "        in_df_fgy = in_df_gy.copy()\n",
    "        in_df_fgy = get_fft(in_df_fgy)\n",
    "\n",
    "        in_df_fgy_jerk = in_df_gy_jerk.copy()\n",
    "        in_df_fgy_jerk = get_fft(in_df_fgy_jerk)\n",
    "\n",
    "        out_gy_df = process_gyr(in_df_gy, tgy_cols, out_gy_df, 0)\n",
    "        out_gy_jerk_df = process_acc(in_df_gy_jerk, tgy_jerk_cols, out_gy_jerk_df, 0)\n",
    "        out_fgy_df = process_acc(in_df_fgy, fgy_cols, out_fgy_df, 1)\n",
    "        out_fgy_jerk_df = process_acc(in_df_fgy_jerk, fgy_jerk_cols, out_fgy_jerk_df, 1)\n",
    "\n",
    "        gy_mag = process_mag(in_df_gy, tgy_mag_cols, gy_mag, 0)\n",
    "        gy_jerk_mag = process_mag(in_df_gy_jerk, tgy_jmag_cols, gy_jerk_mag, 0)\n",
    "        fgy_mag = process_mag(in_df_fgy, fgy_mag_cols, fgy_mag, 1)\n",
    "        fgy_jerk_mag = process_mag(in_df_fgy_jerk, fgy_jmag_cols, fgy_jerk_mag, 1)\n",
    "\n",
    "\n",
    "        #handling mismatch in rows\n",
    "        if out_gy_df.shape[0] < out_acc_df.shape[0]:\n",
    "            out_acc_df = out_acc_df[:out_gy_df.shape[0]]\n",
    "            out_acc_jerk_df = out_acc_jerk_df[:out_gy_df.shape[0]]\n",
    "            out_facc_df = out_facc_df[:out_gy_df.shape[0]]\n",
    "            out_facc_jerk_df = out_facc_jerk_df[:out_gy_df.shape[0]]\n",
    "            acc_mag = acc_mag[:out_gy_df.shape[0]]\n",
    "            acc_jerk_mag = acc_jerk_mag[:out_gy_df.shape[0]]\n",
    "            facc_mag = facc_mag[:out_gy_df.shape[0]]\n",
    "            facc_jerk_mag = facc_jerk_mag[:out_gy_df.shape[0]]\n",
    "\n",
    "        acc_gy_df = pd.concat([out_acc_df, acc_mag, out_acc_jerk_df, acc_jerk_mag, \n",
    "                               out_facc_df, facc_mag, out_facc_jerk_df, facc_jerk_mag,\n",
    "                               out_gy_df, gy_mag, out_gy_jerk_df, gy_jerk_mag, \n",
    "                               out_fgy_df, fgy_mag, out_fgy_jerk_df, fgy_jerk_mag], axis=1)\n",
    "\n",
    "    print(acc_gy_df.shape)\n",
    "    return acc_gy_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|                                                                                           | 0/40 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Tue_Mar_06_18-33_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Tue_Mar_06_18-33_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  2%|██                                                                              | 1/40 [02:28<1:36:23, 148.29s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Wed_Mar_07_11-53_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Wed_Mar_07_11-53_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  5%|████                                                                            | 2/40 [04:27<1:24:34, 133.54s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Wed_Mar_07_13-36_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Wed_Mar_07_13-36_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  8%|██████                                                                          | 3/40 [05:02<1:02:05, 100.69s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Wed_Mar_07_13-42_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Wed_Mar_07_13-42_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 10%|████████▎                                                                          | 4/40 [05:07<46:07, 76.87s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Wed_Mar_07_13-44_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Wed_Mar_07_13-44_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 12%|██████████                                                                      | 5/40 [08:42<1:00:59, 104.57s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Wed_Mar_07_13-57_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Wed_Mar_07_13-57_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 15%|████████████▍                                                                      | 6/40 [08:45<49:35, 87.51s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_00-16_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_00-16_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 18%|██████████████▌                                                                    | 7/40 [08:49<41:37, 75.67s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_00-18_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_00-18_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 20%|████████████████▌                                                                  | 8/40 [08:53<35:34, 66.69s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_00-25_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_00-25_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 22%|██████████████████▋                                                                | 9/40 [08:57<30:51, 59.72s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_00-31_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_00-31_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 25%|████████████████████▌                                                             | 10/40 [09:21<28:03, 56.13s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_00-35_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_00-35_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 28%|██████████████████████▌                                                           | 11/40 [10:07<26:41, 55.21s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_00-38_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_00-38_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 30%|████████████████████████▌                                                         | 12/40 [10:11<23:46, 50.95s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_00-39_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_00-39_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 32%|██████████████████████████▋                                                       | 13/40 [10:40<22:10, 49.27s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_07-43_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_07-43_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 35%|████████████████████████████▋                                                     | 14/40 [14:38<27:12, 62.79s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_07-54_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_07-54_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 38%|██████████████████████████████▊                                                   | 15/40 [19:23<32:18, 77.54s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_08-13_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_08-13_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 40%|████████████████████████████████▊                                                 | 16/40 [26:13<39:19, 98.32s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_14-46_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_14-46_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 42%|██████████████████████████████████▍                                              | 17/40 [37:52<51:14, 133.69s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_19-19_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Thu_Feb_08_19-19_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 45%|████████████████████████████████████▍                                            | 18/40 [39:09<47:51, 130.55s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Fri_Feb_09_09-37_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Fri_Feb_09_09-37_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 48%|██████████████████████████████████████▍                                          | 19/40 [40:08<44:21, 126.76s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Fri_Feb_09_09-41_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Fri_Feb_09_09-41_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 50%|████████████████████████████████████████▌                                        | 20/40 [41:34<41:34, 124.72s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Fri_Feb_09_09-47_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Fri_Feb_09_09-47_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 52%|██████████████████████████████████████████▌                                      | 21/40 [43:45<39:35, 125.04s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Fri_Feb_09_09-54_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Fri_Feb_09_09-54_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 55%|████████████████████████████████████████████▌                                    | 22/40 [49:19<40:21, 134.51s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Fri_Feb_09_11-54_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Fri_Feb_09_11-54_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 57%|██████████████████████████████████████████████▌                                  | 23/40 [54:54<40:35, 143.24s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Sun_Feb_11_18-31_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Sun_Feb_11_18-31_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 60%|███████████████████████████████████████████████▍                               | 24/40 [1:10:02<46:41, 175.12s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Mon_Feb_12_01-46_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Mon_Feb_12_01-46_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 62%|█████████████████████████████████████████████████▍                             | 25/40 [1:21:36<48:57, 195.86s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Mon_Feb_12_22-04_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Mon_Feb_12_22-04_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 65%|███████████████████████████████████████████████████▎                           | 26/40 [1:28:03<47:25, 203.23s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Tue_Feb_13_00-40_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Tue_Feb_13_00-40_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 68%|█████████████████████████████████████████████████████▎                         | 27/40 [1:35:55<46:11, 213.16s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Tue_Feb_13_17-53_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Tue_Feb_13_17-53_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 70%|███████████████████████████████████████████████████████▎                       | 28/40 [1:37:42<41:52, 209.38s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Tue_Feb_13_18-01_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Tue_Feb_13_18-01_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 72%|█████████████████████████████████████████████████████████▎                     | 29/40 [1:40:03<37:57, 207.01s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Tue_Feb_13_18-09_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Tue_Feb_13_18-09_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 75%|███████████████████████████████████████████████████████████▎                   | 30/40 [1:50:26<36:48, 220.89s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Fri_Feb_16_14-11_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Fri_Feb_16_14-11_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 78%|█████████████████████████████████████████████████████████████▏                 | 31/40 [1:53:55<33:04, 220.51s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Fri_Feb_16_14-21_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Fri_Feb_16_14-21_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 80%|███████████████████████████████████████████████████████████████▏               | 32/40 [1:58:21<29:35, 221.91s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Fri_Feb_16_14-33_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Fri_Feb_16_14-33_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 82%|█████████████████████████████████████████████████████████████████▏             | 33/40 [2:04:55<26:29, 227.13s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Fri_Feb_16_17-01_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Fri_Feb_16_17-01_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 85%|███████████████████████████████████████████████████████████████████▏           | 34/40 [2:59:59<31:45, 317.62s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Fri_Feb_16_18-24_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Fri_Feb_16_18-24_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 88%|█████████████████████████████████████████████████████████████████████▏         | 35/40 [3:00:02<25:43, 308.64s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Mon_Feb_26_13-54_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Mon_Feb_26_13-54_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 90%|███████████████████████████████████████████████████████████████████████        | 36/40 [3:00:28<20:03, 300.81s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Mon_Feb_26_15-18_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Mon_Feb_26_15-18_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 92%|█████████████████████████████████████████████████████████████████████████      | 37/40 [3:02:48<14:49, 296.45s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Mon_Feb_26_16-04_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Mon_Feb_26_16-04_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 95%|███████████████████████████████████████████████████████████████████████████    | 38/40 [3:05:33<09:45, 292.99s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Mon_Feb_26_18-01_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Mon_Feb_26_18-01_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 98%|█████████████████████████████████████████████████████████████████████████████  | 39/40 [3:06:01<04:46, 286.20s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: Sessions_new\\14442D57F7E1B88_Mon_Feb_26_18-03_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: Sessions_new\\14442D57F7E1B88_Mon_Feb_26_18-03_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████████████████████████████████████████| 40/40 [3:09:25<00:00, 284.14s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3934, 289)\n",
      "--- 11366.227885007858 seconds ---\n"
     ]
    }
   ],
   "source": [
    "start_time = time.time()\n",
    "training_set_df = process_csv_data(\"Sessions_new/\")\n",
    "print(\"--- %s seconds ---\" % (time.time() - start_time))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "training_set_df.to_csv(\"10sec_smooth_cropped_288feat.csv\")\n",
    "#hist_activities(training_set_df)\n",
    "# plot_act_signals(in_df_ac)\n",
    "\n",
    "#combined_training_df = training_set_df\n",
    "#combined_training_df = combined_training_df.iloc[np.random.permutation(len(combined_training_df))]\n",
    "#combined_training_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "64 288\n",
      "64\n",
      "['walking' 'walking' 'walking' 'walking' 'walking' 'walking' 'walking'\n",
      " 'walking' 'walking' 'walking' 'walking' 'walking' 'walking' 'walking'\n",
      " 'walking' 'walking' 'walking' 'walking' 'walking' 'walking' 'walking'\n",
      " 'walking' 'walking' 'walking' 'walking' 'walking' 'walking' 'walking'\n",
      " 'walking' 'walking' 'walking' 'walking' 'walking' 'walking' 'walking'\n",
      " 'walking' 'walking' 'walking' 'walking' 'walking' 'walking' 'walking'\n",
      " 'walking' 'walking' 'walking' 'walking' 'walking' 'walking' 'walking'\n",
      " 'walking' 'walking' 'walking' 'walking' 'walking' 'walking' 'walking'\n",
      " 'walking' 'walking' 'walking' 'walking' 'walking' 'walking' 'walking'\n",
      " 'walking']\n",
      "(64,)\n",
      "64\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/gringotts/anaconda3/lib/python3.6/site-packages/sklearn/utils/validation.py:475: DataConversionWarning: Data with input dtype object was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[2, 2, 2, 2, 2]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Training set\n",
    "\n",
    "training_data = combined_training_df[tacc_cols+tacc_jerk_cols+tacc_mag_cols+tacc_jmag_cols+facc_cols+facc_jerk_cols+facc_mag_cols+facc_jmag_cols+tgy_jerk_cols+tgy_mag_cols+tgy_jmag_cols+fgy_cols+fgy_jerk_cols+fgy_mag_cols+fgy_jmag_cols+tgy_cols].as_matrix()\n",
    "training_targets = training_data[:, -1]\n",
    "training_data = training_data[:, :-1]\n",
    "\n",
    "std_scale = preprocessing.StandardScaler().fit(training_data)\n",
    "training_data_std = std_scale.transform(training_data)\n",
    "\n",
    "print(len(training_data_std), len(training_data_std[0]))\n",
    "print(len(training_targets))\n",
    "\n",
    "cache = {'laying_down' : 0, 'sitting' : 1, 'walking' : 2, 'standing' : 3}\n",
    "training_targets.reshape(len(training_targets))\n",
    "print(training_targets)\n",
    "print(training_targets.shape)\n",
    "\n",
    "training_bin_targets = []\n",
    "for i in training_targets:\n",
    "    training_bin_targets.append(cache[i])\n",
    "print(len(training_bin_targets))\n",
    "\n",
    "training_bin_targets[: 5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|                                                                                            | 0/4 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: newData1\\14442D53F81BFEA_Mon_Mar_05_21-54_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: newData1\\14442D53F81BFEA_Mon_Mar_05_21-54_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 25%|████████████████████▊                                                              | 1/4 [02:55<08:47, 175.86s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: newData1\\14442D53F81BFEA_Mon_Mar_05_22-01_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: newData1\\14442D53F81BFEA_Mon_Mar_05_22-01_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 50%|█████████████████████████████████████████▌                                         | 2/4 [04:31<04:31, 135.99s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: newData1\\14442D53F81BFEA_Mon_Mar_05_22-06_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: newData1\\14442D53F81BFEA_Mon_Mar_05_22-06_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 75%|██████████████████████████████████████████████████████████████▎                    | 3/4 [06:21<02:07, 127.25s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing acceleration: newData1\\14442D53F81BFEA_Mon_Mar_05_22-14_2018_PST\\data\\1_android.sensor.accelerometer.data.csv.gz\n",
      "processing gyroscope: newData1\\14442D53F81BFEA_Mon_Mar_05_22-14_2018_PST\\data\\4_android.sensor.gyroscope.data.csv.gz\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████████████████████████████████████████████| 4/4 [08:19<00:00, 124.78s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(241, 289)\n"
     ]
    }
   ],
   "source": [
    "## Testing set\n",
    "testing_set_df = process_csv_data('newData1/')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfwAAAG7CAYAAAAmDyfHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x22d00866048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean_acc_x</th>\n",
       "      <th>mean_acc_y</th>\n",
       "      <th>mean_acc_z</th>\n",
       "      <th>std_acc_x</th>\n",
       "      <th>std_acc_y</th>\n",
       "      <th>std_acc_z</th>\n",
       "      <th>mad_acc_x</th>\n",
       "      <th>mad_acc_y</th>\n",
       "      <th>mad_acc_z</th>\n",
       "      <th>min_acc_x</th>\n",
       "      <th>...</th>\n",
       "      <th>std_mfgy_jerk</th>\n",
       "      <th>mad_mfgy_jerk</th>\n",
       "      <th>min_mfgy_jerk</th>\n",
       "      <th>max_mfgy_jerk</th>\n",
       "      <th>em_mfgy_jerk</th>\n",
       "      <th>iqr_mfgy_jerk</th>\n",
       "      <th>sma_mfgy_jerk</th>\n",
       "      <th>skew_mfgy_jerk</th>\n",
       "      <th>kurt_mfgy_jerk</th>\n",
       "      <th>idx_mfgy_jerk</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>2.766752</td>\n",
       "      <td>-5.620212</td>\n",
       "      <td>-4.253272</td>\n",
       "      <td>1.274890</td>\n",
       "      <td>4.001141</td>\n",
       "      <td>4.968234</td>\n",
       "      <td>0.861058</td>\n",
       "      <td>2.405195</td>\n",
       "      <td>2.380841</td>\n",
       "      <td>-0.507961</td>\n",
       "      <td>...</td>\n",
       "      <td>1107.631634</td>\n",
       "      <td>787.089329</td>\n",
       "      <td>369.126376</td>\n",
       "      <td>6304.403006</td>\n",
       "      <td>9.413842e+06</td>\n",
       "      <td>1612.245561</td>\n",
       "      <td>2861.292449</td>\n",
       "      <td>0.311262</td>\n",
       "      <td>2.551548</td>\n",
       "      <td>832.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>192</th>\n",
       "      <td>1.305884</td>\n",
       "      <td>-1.261114</td>\n",
       "      <td>8.433857</td>\n",
       "      <td>2.256954</td>\n",
       "      <td>3.642098</td>\n",
       "      <td>0.563214</td>\n",
       "      <td>0.456413</td>\n",
       "      <td>0.805582</td>\n",
       "      <td>0.290685</td>\n",
       "      <td>-4.297983</td>\n",
       "      <td>...</td>\n",
       "      <td>612.488971</td>\n",
       "      <td>398.853883</td>\n",
       "      <td>460.164187</td>\n",
       "      <td>4612.493720</td>\n",
       "      <td>3.815453e+06</td>\n",
       "      <td>805.165372</td>\n",
       "      <td>1854.807227</td>\n",
       "      <td>0.632278</td>\n",
       "      <td>3.786425</td>\n",
       "      <td>22.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>144</th>\n",
       "      <td>-6.170257</td>\n",
       "      <td>-7.286478</td>\n",
       "      <td>-1.777548</td>\n",
       "      <td>1.468508</td>\n",
       "      <td>2.360041</td>\n",
       "      <td>1.859090</td>\n",
       "      <td>1.004313</td>\n",
       "      <td>1.745025</td>\n",
       "      <td>1.134163</td>\n",
       "      <td>-11.102040</td>\n",
       "      <td>...</td>\n",
       "      <td>1381.192342</td>\n",
       "      <td>877.302842</td>\n",
       "      <td>729.825309</td>\n",
       "      <td>9326.553424</td>\n",
       "      <td>1.562331e+07</td>\n",
       "      <td>1814.065770</td>\n",
       "      <td>3703.459852</td>\n",
       "      <td>0.759168</td>\n",
       "      <td>3.450777</td>\n",
       "      <td>180.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>3.862147</td>\n",
       "      <td>-8.913618</td>\n",
       "      <td>-1.631911</td>\n",
       "      <td>0.130231</td>\n",
       "      <td>2.083705</td>\n",
       "      <td>0.215592</td>\n",
       "      <td>0.041440</td>\n",
       "      <td>0.022318</td>\n",
       "      <td>0.050673</td>\n",
       "      <td>3.217985</td>\n",
       "      <td>...</td>\n",
       "      <td>1159.297859</td>\n",
       "      <td>736.572607</td>\n",
       "      <td>805.457656</td>\n",
       "      <td>8481.740311</td>\n",
       "      <td>1.158101e+07</td>\n",
       "      <td>1508.542646</td>\n",
       "      <td>3199.537699</td>\n",
       "      <td>0.772122</td>\n",
       "      <td>3.713802</td>\n",
       "      <td>136.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106</th>\n",
       "      <td>-1.646043</td>\n",
       "      <td>-6.470183</td>\n",
       "      <td>-3.640172</td>\n",
       "      <td>5.911502</td>\n",
       "      <td>2.074375</td>\n",
       "      <td>2.474632</td>\n",
       "      <td>2.730464</td>\n",
       "      <td>1.099797</td>\n",
       "      <td>1.554656</td>\n",
       "      <td>-9.487904</td>\n",
       "      <td>...</td>\n",
       "      <td>544.328254</td>\n",
       "      <td>360.292849</td>\n",
       "      <td>323.625419</td>\n",
       "      <td>3721.806916</td>\n",
       "      <td>2.945374e+06</td>\n",
       "      <td>718.941858</td>\n",
       "      <td>1627.599774</td>\n",
       "      <td>0.520601</td>\n",
       "      <td>3.204321</td>\n",
       "      <td>1084.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 289 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     mean_acc_x  mean_acc_y  mean_acc_z  std_acc_x  std_acc_y  std_acc_z  \\\n",
       "33     2.766752   -5.620212   -4.253272   1.274890   4.001141   4.968234   \n",
       "192    1.305884   -1.261114    8.433857   2.256954   3.642098   0.563214   \n",
       "144   -6.170257   -7.286478   -1.777548   1.468508   2.360041   1.859090   \n",
       "11     3.862147   -8.913618   -1.631911   0.130231   2.083705   0.215592   \n",
       "106   -1.646043   -6.470183   -3.640172   5.911502   2.074375   2.474632   \n",
       "\n",
       "     mad_acc_x  mad_acc_y  mad_acc_z  min_acc_x      ...        std_mfgy_jerk  \\\n",
       "33    0.861058   2.405195   2.380841  -0.507961      ...          1107.631634   \n",
       "192   0.456413   0.805582   0.290685  -4.297983      ...           612.488971   \n",
       "144   1.004313   1.745025   1.134163 -11.102040      ...          1381.192342   \n",
       "11    0.041440   0.022318   0.050673   3.217985      ...          1159.297859   \n",
       "106   2.730464   1.099797   1.554656  -9.487904      ...           544.328254   \n",
       "\n",
       "     mad_mfgy_jerk  min_mfgy_jerk  max_mfgy_jerk  em_mfgy_jerk  iqr_mfgy_jerk  \\\n",
       "33      787.089329     369.126376    6304.403006  9.413842e+06    1612.245561   \n",
       "192     398.853883     460.164187    4612.493720  3.815453e+06     805.165372   \n",
       "144     877.302842     729.825309    9326.553424  1.562331e+07    1814.065770   \n",
       "11      736.572607     805.457656    8481.740311  1.158101e+07    1508.542646   \n",
       "106     360.292849     323.625419    3721.806916  2.945374e+06     718.941858   \n",
       "\n",
       "     sma_mfgy_jerk  skew_mfgy_jerk  kurt_mfgy_jerk  idx_mfgy_jerk  \n",
       "33     2861.292449        0.311262        2.551548          832.0  \n",
       "192    1854.807227        0.632278        3.786425           22.0  \n",
       "144    3703.459852        0.759168        3.450777          180.0  \n",
       "11     3199.537699        0.772122        3.713802          136.0  \n",
       "106    1627.599774        0.520601        3.204321         1084.0  \n",
       "\n",
       "[5 rows x 289 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "testing_set_df.to_csv(\"vidushi_288_10sc.csv\")\n",
    "hist_activities(testing_set_df)\n",
    "# plot_act_signals(in_df_ac)\n",
    "\n",
    "combined_testing_df = testing_set_df\n",
    "combined_testing_df = combined_testing_df.iloc[np.random.permutation(len(combined_testing_df))]\n",
    "combined_testing_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "64 288\n",
      "64\n",
      "(64,)\n",
      "64\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/gringotts/anaconda3/lib/python3.6/site-packages/sklearn/utils/validation.py:475: DataConversionWarning: Data with input dtype object was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[2, 2, 2, 2, 2]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "testing_data = combined_testing_df[tacc_cols+tacc_jerk_cols+tacc_mag_cols+tacc_jmag_cols+facc_cols+facc_jerk_cols+facc_mag_cols+facc_jmag_cols+tgy_jerk_cols+tgy_mag_cols+tgy_jmag_cols+fgy_cols+fgy_jerk_cols+fgy_mag_cols+fgy_jmag_cols+tgy_cols].as_matrix()\n",
    "testing_targets = testing_data[:, -1]\n",
    "testing_data = testing_data[:, :-1]\n",
    "\n",
    "std_scale = preprocessing.StandardScaler().fit(testing_data)\n",
    "testing_data_std = std_scale.transform(testing_data)\n",
    "\n",
    "print(len(testing_data_std), len(testing_data_std[0]))\n",
    "print(len(testing_targets))\n",
    "\n",
    "cache = {'laying_down' : 0, 'sitting' : 1, 'walking' : 2, 'standing' : 3}\n",
    "testing_targets.reshape(len(testing_targets))\n",
    "print(testing_targets.shape)\n",
    "\n",
    "testing_bin_targets = []\n",
    "for i in testing_targets:\n",
    "    testing_bin_targets.append(cache[i])\n",
    "print(len(testing_bin_targets))\n",
    "\n",
    "testing_bin_targets[: 5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "model = DecisionTreeClassifier(random_state=0)\n",
    "\n",
    "# Training the model\n",
    "model.fit(training_data, training_bin_targets)\n",
    "\n",
    "# Testing on the testing data\n",
    "predicted_values = model.predict(testing_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score:  1.0\n"
     ]
    }
   ],
   "source": [
    "acc_score = accuracy_score(testing_bin_targets, predicted_values)\n",
    "print(\"Score: \", acc_score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [default]",
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}