Mood_recognition.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [],
   "source": [
    "import librosa\n",
    "import librosa.display\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib.pyplot import specgram\n",
    "from tensorflow.keras.preprocessing import sequence\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense\n",
    "from tensorflow.keras.utils import to_categorical\n",
    "from tensorflow.keras.layers import Input, Flatten, Dropout, Activation\n",
    "from tensorflow.keras.layers import Conv2D, MaxPooling2D, AveragePooling1D\n",
    "from tensorflow.keras.models import Model\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from tensorflow.keras import regularizers\n",
    "import os\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "list"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "mylist= os.listdir('C:\\\\Users\\Kalelt\\'has\\Desktop\\Stage\\Sound Data\\Raw Data')\n",
    "type(mylist)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2932"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(mylist)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "feeling_list=[]\n",
    "for item in mylist:\n",
    "    if item[6:-16]=='03':\n",
    "        feeling_list.append('happy')\n",
    "    elif item[6:-16]=='04':\n",
    "        feeling_list.append('sad')\n",
    "    elif item[6:-16]=='05':\n",
    "        feeling_list.append('angry')\n",
    "    elif item[6:-16]=='06':\n",
    "        feeling_list.append('fearful')\n",
    "    elif item[6:-16]=='07':\n",
    "        feeling_list.append('disgust')\n",
    "    elif item[6:-16]=='08':\n",
    "        feeling_list.append('surprised')\n",
    "    elif item[6:-16]=='01':\n",
    "        feeling_list.append('neutral')\n",
    "    elif item[:1]=='a':\n",
    "        feeling_list.append('angry')\n",
    "    elif item[:1]=='f':\n",
    "        feeling_list.append('fearful')\n",
    "    elif item[:1]=='h':\n",
    "        feeling_list.append('happy')\n",
    "    elif item[:1]=='n':\n",
    "        feeling_list.append('neutral')\n",
    "    elif item[:1]=='d':\n",
    "        feeling_list.append('disgust')\n",
    "    elif item[:2]=='su':\n",
    "        feeling_list.append('surprised')\n",
    "    elif item[:2]=='sa':\n",
    "        feeling_list.append('sad')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2556"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels = pd.DataFrame(feeling_list)\n",
    "labels = labels.rename(columns={0: 'label'})\n",
    "len(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "# Plotting the emotion distribution\n",
    "\n",
    "def plot_emotion_dist(dist, color_code='#C2185B', title=\"Plot\"):\n",
    "    \"\"\"\n",
    "    To plot the data distributioin by class.\n",
    "    Arg:\n",
    "      dist: pandas series of label count. \n",
    "    \"\"\"\n",
    "    tmp_df = pd.DataFrame()\n",
    "    tmp_df['Emotion'] = list(dist.keys())\n",
    "    tmp_df['Count'] = list(dist)\n",
    "    fig, ax = plt.subplots(figsize=(14, 7))\n",
    "    ax = sns.barplot(x=\"Emotion\", y='Count', color=color_code, data=tmp_df)\n",
    "    ax.set_title(title)\n",
    "    ax.set_xticklabels(ax.get_xticklabels(),rotation=45)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "angry        436\n",
      "fearful      436\n",
      "sad          436\n",
      "happy        436\n",
      "neutral      308\n",
      "disgust      252\n",
      "surprised    252\n",
      "Name: label, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot = labels.label.value_counts()\n",
    "print(plot)\n",
    "plot_emotion_dist(plot, \"#2962FF\", \"Emotion Distribution\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0         neutral\n",
       "1         neutral\n",
       "2         neutral\n",
       "3         neutral\n",
       "4         neutral\n",
       "          ...    \n",
       "2551    surprised\n",
       "2552    surprised\n",
       "2553    surprised\n",
       "2554    surprised\n",
       "2555    surprised\n",
       "Name: label, Length: 2556, dtype: object"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels.label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "#Prepare the dict\n",
    "my_dict = {}\n",
    "i=0\n",
    "for i in range(216):\n",
    "    my_dict[i] = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "for index,y in enumerate(mylist):\n",
    "    if mylist[index][6:-16]!='02':\n",
    "        X, sample_rate = librosa.load('C:\\\\Users\\Kalelt\\'has\\Desktop\\Stage\\Sound Data\\Raw Data\\\\'+y, res_type='kaiser_fast',duration=2.5,sr=22050*2,offset=0.5)\n",
    "        sample_rate = np.array(sample_rate)\n",
    "        mfccs = librosa.feature.mfcc(y=X,sr=sample_rate,n_mfcc=12).T\n",
    "        for i in range(len(mfccs)):\n",
    "            my_dict[i].append(mfccs[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "df = pd.DataFrame.from_dict(my_dict, orient='index').T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "xd = df\n",
    "for i in range(len(df)):\n",
    "    for j in range(216):\n",
    "        if xd.iloc[i,j] is None:\n",
    "            xd.iloc[i,j] = np.zeros(12)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>207</th>\n",
       "      <th>208</th>\n",
       "      <th>209</th>\n",
       "      <th>210</th>\n",
       "      <th>211</th>\n",
       "      <th>212</th>\n",
       "      <th>213</th>\n",
       "      <th>214</th>\n",
       "      <th>215</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>[-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>...</td>\n",
       "      <td>[-852.7659, 3.7398262, 3.7303648, 3.7146397, 3...</td>\n",
       "      <td>[-854.7548, 0.9293566, 0.9270228, 0.9231334, 0...</td>\n",
       "      <td>[-855.0615, 0.49615216, 0.49515492, 0.49349436...</td>\n",
       "      <td>[-854.1553, 1.7768778, 1.7735907, 1.7681226, 1...</td>\n",
       "      <td>[-854.2701, 1.6132803, 1.6062467, 1.5947204, 1...</td>\n",
       "      <td>[-854.79333, 0.8751819, 0.8735771, 0.87091243,...</td>\n",
       "      <td>[-853.399, 2.8447864, 2.8360765, 2.8216248, 2....</td>\n",
       "      <td>[-854.20074, 1.7127061, 1.7094843, 1.7041247, ...</td>\n",
       "      <td>[-853.0279, 3.359327, 3.3201904, 3.2559023, 3....</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>[-821.7846, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-821.7846, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-821.7846, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-821.7846, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-821.7846, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-821.13354, -0.45227766, -0.4762841, 0.919759...</td>\n",
       "      <td>[-812.50635, -1.3923113, -2.837748, 10.556901,...</td>\n",
       "      <td>[-812.2545, 0.47208667, -1.3423647, 9.645514, ...</td>\n",
       "      <td>[-820.55914, 0.5437228, 0.2414756, 1.0925915, ...</td>\n",
       "      <td>[-821.62164, -0.108744904, -0.12704813, 0.2273...</td>\n",
       "      <td>...</td>\n",
       "      <td>[-804.8696, 23.238768, 21.388206, 18.740921, 1...</td>\n",
       "      <td>[-802.52905, 25.882265, 22.24023, 17.351395, 1...</td>\n",
       "      <td>[-802.88104, 25.116793, 20.99273, 15.970868, 1...</td>\n",
       "      <td>[-805.04504, 22.240238, 18.569805, 14.087547, ...</td>\n",
       "      <td>[-804.15796, 22.135447, 15.313034, 7.9197245, ...</td>\n",
       "      <td>[-804.4919, 22.181942, 16.381016, 9.514078, 4....</td>\n",
       "      <td>[-805.3632, 21.638456, 17.417744, 11.918913, 6...</td>\n",
       "      <td>[-803.8157, 24.056448, 20.318403, 15.074839, 9...</td>\n",
       "      <td>[-810.08716, 15.698001, 13.357195, 10.0368, 6....</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>[-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>[-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
       "      <td>...</td>\n",
       "      <td>[-768.4601, 19.58313, 16.457134, 14.096729, 13...</td>\n",
       "      <td>[-766.63574, 22.676502, 20.294708, 17.63641, 1...</td>\n",
       "      <td>[-761.7448, 29.168606, 25.644545, 21.469763, 1...</td>\n",
       "      <td>[-763.19244, 27.045345, 23.24071, 18.607624, 1...</td>\n",
       "      <td>[-770.62866, 17.464836, 15.928329, 13.6533165,...</td>\n",
       "      <td>[-772.97015, 14.067352, 12.330967, 9.887375, 7...</td>\n",
       "      <td>[-766.97754, 21.83543, 18.301792, 13.719212, 9...</td>\n",
       "      <td>[-764.3447, 25.480705, 21.830109, 17.265268, 1...</td>\n",
       "      <td>[-765.27563, 24.268303, 21.179089, 17.718077, ...</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>[-864.34515, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...</td>\n",
       "      <td>[-864.34515, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...</td>\n",
       "      <td>[-864.34515, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...</td>\n",
       "      <td>[-864.34515, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...</td>\n",
       "      <td>[-864.34515, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...</td>\n",
       "      <td>[-864.34515, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...</td>\n",
       "      <td>[-860.83636, -4.710763, 3.9831913, -2.8565996,...</td>\n",
       "      <td>[-856.79315, -10.1305, 8.542254, -6.089765, 3....</td>\n",
       "      <td>[-860.0415, -5.7772565, 4.8832717, -3.4995217,...</td>\n",
       "      <td>[-864.34515, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...</td>\n",
       "      <td>...</td>\n",
       "      <td>[-853.5374, 14.629644, 12.938563, 10.728258, 8...</td>\n",
       "      <td>[-856.2662, 10.81254, 9.174643, 6.937224, 4.44...</td>\n",
       "      <td>[-860.00793, 5.4010243, 3.4921875, 1.1485866, ...</td>\n",
       "      <td>[-861.66724, 3.5617552, 3.0976498, 2.7365787, ...</td>\n",
       "      <td>[-857.03516, 8.817564, 5.746298, 3.6268888, 2....</td>\n",
       "      <td>[-855.90045, 11.213295, 9.260658, 6.647816, 3....</td>\n",
       "      <td>[-855.6426, 11.319319, 8.957191, 6.4642467, 4....</td>\n",
       "      <td>[-860.6173, 5.0931444, 4.592247, 3.8604326, 3....</td>\n",
       "      <td>[-864.34515, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>[-864.3776, 8.127356, 7.884954, 7.6639547, 7.5...</td>\n",
       "      <td>[-864.946, 7.3574443, 7.204069, 7.1013308, 7.0...</td>\n",
       "      <td>[-868.55853, 2.3213463, 2.3170085, 2.3097866, ...</td>\n",
       "      <td>[-869.71155, 0.6920187, 0.69145864, 0.6905184,...</td>\n",
       "      <td>[-865.8574, 6.081609, 5.965397, 5.9163666, 5.9...</td>\n",
       "      <td>[-864.7549, 7.320035, 6.6085896, 6.365541, 6.8...</td>\n",
       "      <td>[-866.9905, 4.5374575, 4.529083, 4.515147, 4.4...</td>\n",
       "      <td>[-867.3611, 4.0134726, 4.0053864, 3.991928, 3....</td>\n",
       "      <td>[-865.1668, 7.0757146, 6.984665, 6.9181204, 6....</td>\n",
       "      <td>[-863.0326, 9.387402, 7.8706837, 6.969913, 7.2...</td>\n",
       "      <td>...</td>\n",
       "      <td>[-837.63605, 41.153885, 29.524885, 17.727272, ...</td>\n",
       "      <td>[-832.2829, 46.128418, 29.561947, 15.728878, 1...</td>\n",
       "      <td>[-835.6623, 42.727295, 28.810658, 16.206654, 1...</td>\n",
       "      <td>[-845.30383, 32.340084, 25.161396, 16.834219, ...</td>\n",
       "      <td>[-849.93097, 26.564222, 21.212013, 14.778395, ...</td>\n",
       "      <td>[-846.4228, 30.24308, 22.44601, 15.062183, 11....</td>\n",
       "      <td>[-847.0545, 29.797184, 22.969273, 16.333324, 1...</td>\n",
       "      <td>[-858.9753, 14.669733, 11.819754, 8.911891, 6....</td>\n",
       "      <td>[-867.5775, 3.6292028, 3.400793, 3.0636692, 2....</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 217 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   0  \\\n",
       "0  [-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "1  [-821.7846, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "2  [-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "3  [-864.34515, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...   \n",
       "4  [-864.3776, 8.127356, 7.884954, 7.6639547, 7.5...   \n",
       "\n",
       "                                                   1  \\\n",
       "0  [-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "1  [-821.7846, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "2  [-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "3  [-864.34515, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...   \n",
       "4  [-864.946, 7.3574443, 7.204069, 7.1013308, 7.0...   \n",
       "\n",
       "                                                   2  \\\n",
       "0  [-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "1  [-821.7846, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "2  [-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "3  [-864.34515, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...   \n",
       "4  [-868.55853, 2.3213463, 2.3170085, 2.3097866, ...   \n",
       "\n",
       "                                                   3  \\\n",
       "0  [-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "1  [-821.7846, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "2  [-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "3  [-864.34515, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...   \n",
       "4  [-869.71155, 0.6920187, 0.69145864, 0.6905184,...   \n",
       "\n",
       "                                                   4  \\\n",
       "0  [-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "1  [-821.7846, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "2  [-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "3  [-864.34515, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...   \n",
       "4  [-865.8574, 6.081609, 5.965397, 5.9163666, 5.9...   \n",
       "\n",
       "                                                   5  \\\n",
       "0  [-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "1  [-821.13354, -0.45227766, -0.4762841, 0.919759...   \n",
       "2  [-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "3  [-864.34515, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...   \n",
       "4  [-864.7549, 7.320035, 6.6085896, 6.365541, 6.8...   \n",
       "\n",
       "                                                   6  \\\n",
       "0  [-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "1  [-812.50635, -1.3923113, -2.837748, 10.556901,...   \n",
       "2  [-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "3  [-860.83636, -4.710763, 3.9831913, -2.8565996,...   \n",
       "4  [-866.9905, 4.5374575, 4.529083, 4.515147, 4.4...   \n",
       "\n",
       "                                                   7  \\\n",
       "0  [-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "1  [-812.2545, 0.47208667, -1.3423647, 9.645514, ...   \n",
       "2  [-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "3  [-856.79315, -10.1305, 8.542254, -6.089765, 3....   \n",
       "4  [-867.3611, 4.0134726, 4.0053864, 3.991928, 3....   \n",
       "\n",
       "                                                   8  \\\n",
       "0  [-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "1  [-820.55914, 0.5437228, 0.2414756, 1.0925915, ...   \n",
       "2  [-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...   \n",
       "3  [-860.0415, -5.7772565, 4.8832717, -3.4995217,...   \n",
       "4  [-865.1668, 7.0757146, 6.984665, 6.9181204, 6....   \n",
       "\n",
       "                                                   9  ...  \\\n",
       "0  [-855.4126, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...  ...   \n",
       "1  [-821.62164, -0.108744904, -0.12704813, 0.2273...  ...   \n",
       "2  [-783.3624, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...  ...   \n",
       "3  [-864.34515, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...  ...   \n",
       "4  [-863.0326, 9.387402, 7.8706837, 6.969913, 7.2...  ...   \n",
       "\n",
       "                                                 207  \\\n",
       "0  [-852.7659, 3.7398262, 3.7303648, 3.7146397, 3...   \n",
       "1  [-804.8696, 23.238768, 21.388206, 18.740921, 1...   \n",
       "2  [-768.4601, 19.58313, 16.457134, 14.096729, 13...   \n",
       "3  [-853.5374, 14.629644, 12.938563, 10.728258, 8...   \n",
       "4  [-837.63605, 41.153885, 29.524885, 17.727272, ...   \n",
       "\n",
       "                                                 208  \\\n",
       "0  [-854.7548, 0.9293566, 0.9270228, 0.9231334, 0...   \n",
       "1  [-802.52905, 25.882265, 22.24023, 17.351395, 1...   \n",
       "2  [-766.63574, 22.676502, 20.294708, 17.63641, 1...   \n",
       "3  [-856.2662, 10.81254, 9.174643, 6.937224, 4.44...   \n",
       "4  [-832.2829, 46.128418, 29.561947, 15.728878, 1...   \n",
       "\n",
       "                                                 209  \\\n",
       "0  [-855.0615, 0.49615216, 0.49515492, 0.49349436...   \n",
       "1  [-802.88104, 25.116793, 20.99273, 15.970868, 1...   \n",
       "2  [-761.7448, 29.168606, 25.644545, 21.469763, 1...   \n",
       "3  [-860.00793, 5.4010243, 3.4921875, 1.1485866, ...   \n",
       "4  [-835.6623, 42.727295, 28.810658, 16.206654, 1...   \n",
       "\n",
       "                                                 210  \\\n",
       "0  [-854.1553, 1.7768778, 1.7735907, 1.7681226, 1...   \n",
       "1  [-805.04504, 22.240238, 18.569805, 14.087547, ...   \n",
       "2  [-763.19244, 27.045345, 23.24071, 18.607624, 1...   \n",
       "3  [-861.66724, 3.5617552, 3.0976498, 2.7365787, ...   \n",
       "4  [-845.30383, 32.340084, 25.161396, 16.834219, ...   \n",
       "\n",
       "                                                 211  \\\n",
       "0  [-854.2701, 1.6132803, 1.6062467, 1.5947204, 1...   \n",
       "1  [-804.15796, 22.135447, 15.313034, 7.9197245, ...   \n",
       "2  [-770.62866, 17.464836, 15.928329, 13.6533165,...   \n",
       "3  [-857.03516, 8.817564, 5.746298, 3.6268888, 2....   \n",
       "4  [-849.93097, 26.564222, 21.212013, 14.778395, ...   \n",
       "\n",
       "                                                 212  \\\n",
       "0  [-854.79333, 0.8751819, 0.8735771, 0.87091243,...   \n",
       "1  [-804.4919, 22.181942, 16.381016, 9.514078, 4....   \n",
       "2  [-772.97015, 14.067352, 12.330967, 9.887375, 7...   \n",
       "3  [-855.90045, 11.213295, 9.260658, 6.647816, 3....   \n",
       "4  [-846.4228, 30.24308, 22.44601, 15.062183, 11....   \n",
       "\n",
       "                                                 213  \\\n",
       "0  [-853.399, 2.8447864, 2.8360765, 2.8216248, 2....   \n",
       "1  [-805.3632, 21.638456, 17.417744, 11.918913, 6...   \n",
       "2  [-766.97754, 21.83543, 18.301792, 13.719212, 9...   \n",
       "3  [-855.6426, 11.319319, 8.957191, 6.4642467, 4....   \n",
       "4  [-847.0545, 29.797184, 22.969273, 16.333324, 1...   \n",
       "\n",
       "                                                 214  \\\n",
       "0  [-854.20074, 1.7127061, 1.7094843, 1.7041247, ...   \n",
       "1  [-803.8157, 24.056448, 20.318403, 15.074839, 9...   \n",
       "2  [-764.3447, 25.480705, 21.830109, 17.265268, 1...   \n",
       "3  [-860.6173, 5.0931444, 4.592247, 3.8604326, 3....   \n",
       "4  [-858.9753, 14.669733, 11.819754, 8.911891, 6....   \n",
       "\n",
       "                                                 215    label  \n",
       "0  [-853.0279, 3.359327, 3.3201904, 3.2559023, 3....  neutral  \n",
       "1  [-810.08716, 15.698001, 13.357195, 10.0368, 6....  neutral  \n",
       "2  [-765.27563, 24.268303, 21.179089, 17.718077, ...  neutral  \n",
       "3  [-864.34515, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0...  neutral  \n",
       "4  [-867.5775, 3.6292028, 3.400793, 3.0636692, 2....  neutral  \n",
       "\n",
       "[5 rows x 217 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "newdf = pd.concat([xd,labels], axis=1)\n",
    "newdf[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "pycharm": {
     "is_executing": true
    },
    "scrolled": true
   },
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>207</th>\n",
       "      <th>208</th>\n",
       "      <th>209</th>\n",
       "      <th>210</th>\n",
       "      <th>211</th>\n",
       "      <th>212</th>\n",
       "      <th>213</th>\n",
       "      <th>214</th>\n",
       "      <th>215</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2521</th>\n",
       "      <td>[-496.76712, 31.271992, 27.198406, 21.494696, ...</td>\n",
       "      <td>[-492.37686, 37.228104, 32.43861, 25.682755, 1...</td>\n",
       "      <td>[-489.60153, 40.698624, 34.740936, 26.58688, 1...</td>\n",
       "      <td>[-491.94873, 37.578514, 32.061386, 24.23109, 1...</td>\n",
       "      <td>[-490.73798, 39.39726, 34.172623, 26.736, 18.5...</td>\n",
       "      <td>[-492.5524, 36.842293, 31.738192, 24.716717, 1...</td>\n",
       "      <td>[-492.7633, 36.649918, 31.77442, 24.910816, 17...</td>\n",
       "      <td>[-490.26892, 39.890995, 34.249504, 26.38288, 1...</td>\n",
       "      <td>[-490.146, 40.153194, 34.774258, 27.307104, 19...</td>\n",
       "      <td>[-493.20538, 36.06829, 31.307842, 24.59076, 17...</td>\n",
       "      <td>...</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "      <td>surprised</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>776</th>\n",
       "      <td>[-713.93286, 61.049072, 13.8733635, 29.579771,...</td>\n",
       "      <td>[-722.0853, 59.927254, 17.149971, 24.275846, 3...</td>\n",
       "      <td>[-735.17065, 50.955055, 14.183012, 17.443724, ...</td>\n",
       "      <td>[-741.48425, 48.129997, 16.449556, 15.31377, 2...</td>\n",
       "      <td>[-727.93604, 53.29362, 7.999213, 13.963708, 28...</td>\n",
       "      <td>[-718.64905, 56.576496, 1.8556322, 13.412949, ...</td>\n",
       "      <td>[-714.5063, 58.677437, -2.7600968, 10.569329, ...</td>\n",
       "      <td>[-726.43024, 50.676247, -0.4593429, 12.337333,...</td>\n",
       "      <td>[-748.3313, 41.65054, 13.600557, 14.876021, 29...</td>\n",
       "      <td>[-765.7199, 30.452763, 18.811678, 15.219877, 1...</td>\n",
       "      <td>...</td>\n",
       "      <td>[-689.41095, 86.58216, 5.171112, -3.117579, 17...</td>\n",
       "      <td>[-709.6174, 74.40654, 10.773499, 0.69532037, 1...</td>\n",
       "      <td>[-733.65485, 56.745834, 17.080437, 9.706935, 1...</td>\n",
       "      <td>[-748.5231, 48.315636, 27.961443, 22.159208, 2...</td>\n",
       "      <td>[-753.17456, 45.72464, 29.306944, 19.904867, 1...</td>\n",
       "      <td>[-758.8856, 39.942436, 27.982845, 22.0816, 20....</td>\n",
       "      <td>[-765.70935, 33.968616, 27.35603, 22.106358, 1...</td>\n",
       "      <td>[-767.41077, 31.83524, 26.238026, 22.378586, 2...</td>\n",
       "      <td>[-767.10156, 32.780014, 28.204529, 24.301315, ...</td>\n",
       "      <td>fearful</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1349</th>\n",
       "      <td>[-587.5345, -2.8720486, -12.558419, 7.08062, 7...</td>\n",
       "      <td>[-587.5081, -2.22676, -11.945736, 7.4478216, 8...</td>\n",
       "      <td>[-586.63257, -3.1284013, -13.62468, 8.362898, ...</td>\n",
       "      <td>[-586.63947, -3.5335155, -13.557854, 8.711845,...</td>\n",
       "      <td>[-588.8311, -1.9487345, -10.250934, 6.952609, ...</td>\n",
       "      <td>[-591.3227, -1.1529701, -7.266143, 4.9413843, ...</td>\n",
       "      <td>[-592.5418, -0.23216408, -4.8418484, 4.2452307...</td>\n",
       "      <td>[-590.3586, 0.07090867, -5.4192934, 7.0970173,...</td>\n",
       "      <td>[-589.0062, -1.9018729, -9.228389, 7.56669, 6....</td>\n",
       "      <td>[-587.0895, -1.702589, -9.336509, 9.844554, 6....</td>\n",
       "      <td>...</td>\n",
       "      <td>[-274.27777, 109.834854, -58.95273, -18.723331...</td>\n",
       "      <td>[-282.97998, 88.9328, -53.705456, -2.5085034, ...</td>\n",
       "      <td>[-292.87445, 83.338585, -55.394497, 4.9784904,...</td>\n",
       "      <td>[-293.31686, 85.88193, -65.561386, 4.960014, 1...</td>\n",
       "      <td>[-297.69852, 88.15161, -76.35922, 1.1501765, 8...</td>\n",
       "      <td>[-305.10852, 94.84383, -76.066025, 2.9670506, ...</td>\n",
       "      <td>[-524.4871, 172.19736, 47.50946, 54.71444, -24...</td>\n",
       "      <td>[-497.3349, 200.0136, 59.785065, 57.4128, -24....</td>\n",
       "      <td>[-465.64667, 220.94331, 60.447, 49.684357, -21...</td>\n",
       "      <td>happy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2340</th>\n",
       "      <td>[-553.399, 83.77173, 34.90102, 34.278057, 33.8...</td>\n",
       "      <td>[-544.3823, 88.45181, 33.110733, 35.17581, 30....</td>\n",
       "      <td>[-493.8493, 136.02924, 50.216537, 50.345795, 4...</td>\n",
       "      <td>[-429.63336, 197.81017, 52.685356, 29.772839, ...</td>\n",
       "      <td>[-410.8479, 217.30917, 51.629612, 18.541967, 5...</td>\n",
       "      <td>[-431.31998, 203.53995, 59.163788, 27.824467, ...</td>\n",
       "      <td>[-482.29074, 168.66092, 73.99961, 37.809586, 4...</td>\n",
       "      <td>[-522.16187, 134.57848, 73.0145, 38.71186, 33....</td>\n",
       "      <td>[-545.13104, 114.2971, 71.32767, 39.904892, 28...</td>\n",
       "      <td>[-563.1671, 95.88721, 65.994675, 40.936058, 28...</td>\n",
       "      <td>...</td>\n",
       "      <td>[-399.9201, 193.56389, 115.62196, 66.939445, 8...</td>\n",
       "      <td>[-519.854, 38.8964, 27.504597, 13.524971, 2.03...</td>\n",
       "      <td>[-536.6558, 50.42916, 44.56945, 36.213245, 26....</td>\n",
       "      <td>[-405.4281, 162.11479, 72.85392, 21.28218, -10...</td>\n",
       "      <td>[-428.26544, 120.96013, 76.569855, 52.0886, 40...</td>\n",
       "      <td>[-432.5252, 117.79469, 77.57428, 53.076313, 39...</td>\n",
       "      <td>[-440.02805, 30.27645, -5.519136, 79.44932, 16...</td>\n",
       "      <td>[-378.0776, 53.20069, -51.025112, 95.29291, 21...</td>\n",
       "      <td>[-638.9368, 164.45021, 57.409584, 28.4477, 45....</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>946</th>\n",
       "      <td>[-775.79047, 0.16605562, 0.16571796, 0.1651557...</td>\n",
       "      <td>[-775.23987, 0.94442827, 0.9433935, 0.94166774...</td>\n",
       "      <td>[-775.5841, 0.4578952, 0.45762995, 0.4571932, ...</td>\n",
       "      <td>[-775.6452, 0.37153184, 0.3712936, 0.37089425,...</td>\n",
       "      <td>[-774.5837, 1.8719475, 1.8693612, 1.8650556, 1...</td>\n",
       "      <td>[-773.02655, 4.0724487, 4.0648403, 4.0521812, ...</td>\n",
       "      <td>[-773.88745, 2.8558264, 2.851233, 2.8435936, 2...</td>\n",
       "      <td>[-773.9211, 2.8084526, 2.804224, 2.7971976, 2....</td>\n",
       "      <td>[-769.07996, 5.043729, 2.4871254, 7.632422, 7....</td>\n",
       "      <td>[-766.48846, 5.278264, -0.033172607, 8.327629,...</td>\n",
       "      <td>...</td>\n",
       "      <td>[-750.5288, 32.63141, 25.174671, 18.46136, 15....</td>\n",
       "      <td>[-751.0828, 32.82063, 27.243275, 21.271542, 17...</td>\n",
       "      <td>[-752.4153, 31.56753, 27.34085, 22.279207, 18....</td>\n",
       "      <td>[-753.3577, 30.208847, 25.932484, 20.876026, 1...</td>\n",
       "      <td>[-753.7973, 29.453403, 24.858517, 19.518364, 1...</td>\n",
       "      <td>[-756.6581, 26.292389, 23.772095, 20.374128, 1...</td>\n",
       "      <td>[-756.00104, 27.275757, 24.82285, 21.272156, 1...</td>\n",
       "      <td>[-758.74457, 23.663635, 21.919254, 19.271774, ...</td>\n",
       "      <td>[-761.0498, 20.485569, 18.989557, 16.75481, 14...</td>\n",
       "      <td>disgust</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 217 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                      0  \\\n",
       "2521  [-496.76712, 31.271992, 27.198406, 21.494696, ...   \n",
       "776   [-713.93286, 61.049072, 13.8733635, 29.579771,...   \n",
       "1349  [-587.5345, -2.8720486, -12.558419, 7.08062, 7...   \n",
       "2340  [-553.399, 83.77173, 34.90102, 34.278057, 33.8...   \n",
       "946   [-775.79047, 0.16605562, 0.16571796, 0.1651557...   \n",
       "\n",
       "                                                      1  \\\n",
       "2521  [-492.37686, 37.228104, 32.43861, 25.682755, 1...   \n",
       "776   [-722.0853, 59.927254, 17.149971, 24.275846, 3...   \n",
       "1349  [-587.5081, -2.22676, -11.945736, 7.4478216, 8...   \n",
       "2340  [-544.3823, 88.45181, 33.110733, 35.17581, 30....   \n",
       "946   [-775.23987, 0.94442827, 0.9433935, 0.94166774...   \n",
       "\n",
       "                                                      2  \\\n",
       "2521  [-489.60153, 40.698624, 34.740936, 26.58688, 1...   \n",
       "776   [-735.17065, 50.955055, 14.183012, 17.443724, ...   \n",
       "1349  [-586.63257, -3.1284013, -13.62468, 8.362898, ...   \n",
       "2340  [-493.8493, 136.02924, 50.216537, 50.345795, 4...   \n",
       "946   [-775.5841, 0.4578952, 0.45762995, 0.4571932, ...   \n",
       "\n",
       "                                                      3  \\\n",
       "2521  [-491.94873, 37.578514, 32.061386, 24.23109, 1...   \n",
       "776   [-741.48425, 48.129997, 16.449556, 15.31377, 2...   \n",
       "1349  [-586.63947, -3.5335155, -13.557854, 8.711845,...   \n",
       "2340  [-429.63336, 197.81017, 52.685356, 29.772839, ...   \n",
       "946   [-775.6452, 0.37153184, 0.3712936, 0.37089425,...   \n",
       "\n",
       "                                                      4  \\\n",
       "2521  [-490.73798, 39.39726, 34.172623, 26.736, 18.5...   \n",
       "776   [-727.93604, 53.29362, 7.999213, 13.963708, 28...   \n",
       "1349  [-588.8311, -1.9487345, -10.250934, 6.952609, ...   \n",
       "2340  [-410.8479, 217.30917, 51.629612, 18.541967, 5...   \n",
       "946   [-774.5837, 1.8719475, 1.8693612, 1.8650556, 1...   \n",
       "\n",
       "                                                      5  \\\n",
       "2521  [-492.5524, 36.842293, 31.738192, 24.716717, 1...   \n",
       "776   [-718.64905, 56.576496, 1.8556322, 13.412949, ...   \n",
       "1349  [-591.3227, -1.1529701, -7.266143, 4.9413843, ...   \n",
       "2340  [-431.31998, 203.53995, 59.163788, 27.824467, ...   \n",
       "946   [-773.02655, 4.0724487, 4.0648403, 4.0521812, ...   \n",
       "\n",
       "                                                      6  \\\n",
       "2521  [-492.7633, 36.649918, 31.77442, 24.910816, 17...   \n",
       "776   [-714.5063, 58.677437, -2.7600968, 10.569329, ...   \n",
       "1349  [-592.5418, -0.23216408, -4.8418484, 4.2452307...   \n",
       "2340  [-482.29074, 168.66092, 73.99961, 37.809586, 4...   \n",
       "946   [-773.88745, 2.8558264, 2.851233, 2.8435936, 2...   \n",
       "\n",
       "                                                      7  \\\n",
       "2521  [-490.26892, 39.890995, 34.249504, 26.38288, 1...   \n",
       "776   [-726.43024, 50.676247, -0.4593429, 12.337333,...   \n",
       "1349  [-590.3586, 0.07090867, -5.4192934, 7.0970173,...   \n",
       "2340  [-522.16187, 134.57848, 73.0145, 38.71186, 33....   \n",
       "946   [-773.9211, 2.8084526, 2.804224, 2.7971976, 2....   \n",
       "\n",
       "                                                      8  \\\n",
       "2521  [-490.146, 40.153194, 34.774258, 27.307104, 19...   \n",
       "776   [-748.3313, 41.65054, 13.600557, 14.876021, 29...   \n",
       "1349  [-589.0062, -1.9018729, -9.228389, 7.56669, 6....   \n",
       "2340  [-545.13104, 114.2971, 71.32767, 39.904892, 28...   \n",
       "946   [-769.07996, 5.043729, 2.4871254, 7.632422, 7....   \n",
       "\n",
       "                                                      9  ...  \\\n",
       "2521  [-493.20538, 36.06829, 31.307842, 24.59076, 17...  ...   \n",
       "776   [-765.7199, 30.452763, 18.811678, 15.219877, 1...  ...   \n",
       "1349  [-587.0895, -1.702589, -9.336509, 9.844554, 6....  ...   \n",
       "2340  [-563.1671, 95.88721, 65.994675, 40.936058, 28...  ...   \n",
       "946   [-766.48846, 5.278264, -0.033172607, 8.327629,...  ...   \n",
       "\n",
       "                                                    207  \\\n",
       "2521  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...   \n",
       "776   [-689.41095, 86.58216, 5.171112, -3.117579, 17...   \n",
       "1349  [-274.27777, 109.834854, -58.95273, -18.723331...   \n",
       "2340  [-399.9201, 193.56389, 115.62196, 66.939445, 8...   \n",
       "946   [-750.5288, 32.63141, 25.174671, 18.46136, 15....   \n",
       "\n",
       "                                                    208  \\\n",
       "2521  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...   \n",
       "776   [-709.6174, 74.40654, 10.773499, 0.69532037, 1...   \n",
       "1349  [-282.97998, 88.9328, -53.705456, -2.5085034, ...   \n",
       "2340  [-519.854, 38.8964, 27.504597, 13.524971, 2.03...   \n",
       "946   [-751.0828, 32.82063, 27.243275, 21.271542, 17...   \n",
       "\n",
       "                                                    209  \\\n",
       "2521  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...   \n",
       "776   [-733.65485, 56.745834, 17.080437, 9.706935, 1...   \n",
       "1349  [-292.87445, 83.338585, -55.394497, 4.9784904,...   \n",
       "2340  [-536.6558, 50.42916, 44.56945, 36.213245, 26....   \n",
       "946   [-752.4153, 31.56753, 27.34085, 22.279207, 18....   \n",
       "\n",
       "                                                    210  \\\n",
       "2521  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...   \n",
       "776   [-748.5231, 48.315636, 27.961443, 22.159208, 2...   \n",
       "1349  [-293.31686, 85.88193, -65.561386, 4.960014, 1...   \n",
       "2340  [-405.4281, 162.11479, 72.85392, 21.28218, -10...   \n",
       "946   [-753.3577, 30.208847, 25.932484, 20.876026, 1...   \n",
       "\n",
       "                                                    211  \\\n",
       "2521  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...   \n",
       "776   [-753.17456, 45.72464, 29.306944, 19.904867, 1...   \n",
       "1349  [-297.69852, 88.15161, -76.35922, 1.1501765, 8...   \n",
       "2340  [-428.26544, 120.96013, 76.569855, 52.0886, 40...   \n",
       "946   [-753.7973, 29.453403, 24.858517, 19.518364, 1...   \n",
       "\n",
       "                                                    212  \\\n",
       "2521  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...   \n",
       "776   [-758.8856, 39.942436, 27.982845, 22.0816, 20....   \n",
       "1349  [-305.10852, 94.84383, -76.066025, 2.9670506, ...   \n",
       "2340  [-432.5252, 117.79469, 77.57428, 53.076313, 39...   \n",
       "946   [-756.6581, 26.292389, 23.772095, 20.374128, 1...   \n",
       "\n",
       "                                                    213  \\\n",
       "2521  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...   \n",
       "776   [-765.70935, 33.968616, 27.35603, 22.106358, 1...   \n",
       "1349  [-524.4871, 172.19736, 47.50946, 54.71444, -24...   \n",
       "2340  [-440.02805, 30.27645, -5.519136, 79.44932, 16...   \n",
       "946   [-756.00104, 27.275757, 24.82285, 21.272156, 1...   \n",
       "\n",
       "                                                    214  \\\n",
       "2521  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...   \n",
       "776   [-767.41077, 31.83524, 26.238026, 22.378586, 2...   \n",
       "1349  [-497.3349, 200.0136, 59.785065, 57.4128, -24....   \n",
       "2340  [-378.0776, 53.20069, -51.025112, 95.29291, 21...   \n",
       "946   [-758.74457, 23.663635, 21.919254, 19.271774, ...   \n",
       "\n",
       "                                                    215      label  \n",
       "2521  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...  surprised  \n",
       "776   [-767.10156, 32.780014, 28.204529, 24.301315, ...    fearful  \n",
       "1349  [-465.64667, 220.94331, 60.447, 49.684357, -21...      happy  \n",
       "2340  [-638.9368, 164.45021, 57.409584, 28.4477, 45....    neutral  \n",
       "946   [-761.0498, 20.485569, 18.989557, 16.75481, 14...    disgust  \n",
       "\n",
       "[5 rows x 217 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.utils import shuffle\n",
    "rnewdf = shuffle(newdf)\n",
    "rnewdf[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2041"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "newdf1 = np.random.rand(len(rnewdf)) < 0.8\n",
    "train = rnewdf[newdf1]\n",
    "test = rnewdf[~newdf1]\n",
    "len(train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "train_features = train.iloc[:,:-1]\n",
    "train_labels = train.iloc[:, -1:]\n",
    "test_features = test.iloc[:,:-1]\n",
    "test_labels = test.iloc[:, -1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fearful      93\n",
      "happy        90\n",
      "sad          82\n",
      "angry        77\n",
      "neutral      59\n",
      "disgust      59\n",
      "surprised    55\n",
      "Name: label, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot = test_labels.label.value_counts()\n",
    "print(plot)\n",
    "plot_emotion_dist(plot, \"#2962FF\", \"Emotion Distribution\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2041, 216, 12, 1)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Reshaping train and test into 4d array\n",
    "X_train = np.array(train_features)\n",
    "X_train_cnn = np.zeros((len(X_train),216,12))\n",
    "for i in range(len(X_train)):\n",
    "    for j in range(216):\n",
    "        for k in range(12):\n",
    "            X_train_cnn[i,j,k] = X_train[i][j][k]\n",
    "            \n",
    "X_train_cnn = X_train_cnn.reshape(len(train_features),216,12,1)            \n",
    "X_train_cnn.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "#Normalize X train\n",
    "Max = np.amax(X_train_cnn)\n",
    "Min = np.amin(X_train_cnn)\n",
    "X_train_cnn = (X_train_cnn - Min) / (Max-Min)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(515, 216, 12, 1)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test = np.array(test_features)\n",
    "X_test_cnn = np.zeros((len(X_test),216,12))\n",
    "for i in range(len(X_test)):\n",
    "    for j in range(216):\n",
    "        for k in range(12):\n",
    "            X_test_cnn[i,j,k] = X_test[i][j][k]\n",
    "X_test_cnn = X_test_cnn.reshape(len(test_features),216,12,1)\n",
    "X_test_cnn.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "#Normalize X test\n",
    "Max = np.amax(X_test_cnn)\n",
    "Min = np.amin(X_test_cnn)\n",
    "X_test_cnn = (X_test_cnn - Min) / (Max-Min)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "numpy.ndarray"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train = np.array(train_labels)\n",
    "y_test = np.array(test_labels)\n",
    "type(y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Save np array to file\n",
    "np.savetxt(\"lb_feature\", y_train, fmt='%s')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([['sad'],\n",
       "       ['neutral'],\n",
       "       ['surprised'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['sad'],\n",
       "       ['surprised'],\n",
       "       ['angry'],\n",
       "       ['angry'],\n",
       "       ['surprised'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['angry'],\n",
       "       ['surprised'],\n",
       "       ['happy'],\n",
       "       ['sad'],\n",
       "       ['sad'],\n",
       "       ['angry'],\n",
       "       ['surprised'],\n",
       "       ['neutral'],\n",
       "       ['surprised'],\n",
       "       ['neutral'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['disgust'],\n",
       "       ['angry'],\n",
       "       ['surprised'],\n",
       "       ['happy'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['neutral'],\n",
       "       ['angry'],\n",
       "       ['disgust'],\n",
       "       ['angry'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['fearful'],\n",
       "       ['sad'],\n",
       "       ['surprised'],\n",
       "       ['happy'],\n",
       "       ['angry'],\n",
       "       ['happy'],\n",
       "       ['surprised'],\n",
       "       ['angry'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['neutral'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['surprised'],\n",
       "       ['neutral'],\n",
       "       ['angry'],\n",
       "       ['happy'],\n",
       "       ['surprised'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['neutral'],\n",
       "       ['surprised'],\n",
       "       ['neutral'],\n",
       "       ['fearful'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['surprised'],\n",
       "       ['fearful'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['angry'],\n",
       "       ['sad'],\n",
       "       ['disgust'],\n",
       "       ['sad'],\n",
       "       ['disgust'],\n",
       "       ['angry'],\n",
       "       ['happy'],\n",
       "       ['angry'],\n",
       "       ['neutral'],\n",
       "       ['sad'],\n",
       "       ['surprised'],\n",
       "       ['surprised'],\n",
       "       ['neutral'],\n",
       "       ['surprised'],\n",
       "       ['angry'],\n",
       "       ['disgust'],\n",
       "       ['angry'],\n",
       "       ['angry'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['fearful'],\n",
       "       ['happy'],\n",
       "       ['surprised'],\n",
       "       ['sad'],\n",
       "       ['sad'],\n",
       "       ['surprised'],\n",
       "       ['fearful'],\n",
       "       ['sad'],\n",
       "       ['disgust'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['happy'],\n",
       "       ['sad'],\n",
       "       ['neutral'],\n",
       "       ['angry'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['angry'],\n",
       "       ['neutral'],\n",
       "       ['happy'],\n",
       "       ['disgust'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['disgust'],\n",
       "       ['neutral'],\n",
       "       ['sad'],\n",
       "       ['angry'],\n",
       "       ['angry'],\n",
       "       ['happy'],\n",
       "       ['disgust'],\n",
       "       ['disgust'],\n",
       "       ['neutral'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['surprised'],\n",
       "       ['fearful'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['neutral'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['disgust'],\n",
       "       ['neutral'],\n",
       "       ['happy'],\n",
       "       ['angry'],\n",
       "       ['sad'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['surprised'],\n",
       "       ['sad'],\n",
       "       ['fearful'],\n",
       "       ['surprised'],\n",
       "       ['sad'],\n",
       "       ['fearful'],\n",
       "       ['sad'],\n",
       "       ['sad'],\n",
       "       ['surprised'],\n",
       "       ['sad'],\n",
       "       ['fearful'],\n",
       "       ['surprised'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['disgust'],\n",
       "       ['angry'],\n",
       "       ['sad'],\n",
       "       ['surprised'],\n",
       "       ['happy'],\n",
       "       ['disgust'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['neutral'],\n",
       "       ['surprised'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['fearful'],\n",
       "       ['sad'],\n",
       "       ['fearful'],\n",
       "       ['sad'],\n",
       "       ['sad'],\n",
       "       ['disgust'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['surprised'],\n",
       "       ['disgust'],\n",
       "       ['surprised'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['disgust'],\n",
       "       ['sad'],\n",
       "       ['surprised'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['sad'],\n",
       "       ['fearful'],\n",
       "       ['angry'],\n",
       "       ['neutral'],\n",
       "       ['angry'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['angry'],\n",
       "       ['neutral'],\n",
       "       ['fearful'],\n",
       "       ['happy'],\n",
       "       ['sad'],\n",
       "       ['fearful'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['neutral'],\n",
       "       ['happy'],\n",
       "       ['surprised'],\n",
       "       ['angry'],\n",
       "       ['angry'],\n",
       "       ['surprised'],\n",
       "       ['fearful'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['happy'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['surprised'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['sad'],\n",
       "       ['sad'],\n",
       "       ['disgust'],\n",
       "       ['neutral'],\n",
       "       ['neutral'],\n",
       "       ['neutral'],\n",
       "       ['sad'],\n",
       "       ['fearful'],\n",
       "       ['sad'],\n",
       "       ['sad'],\n",
       "       ['sad'],\n",
       "       ['surprised'],\n",
       "       ['neutral'],\n",
       "       ['happy'],\n",
       "       ['disgust'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['neutral'],\n",
       "       ['neutral'],\n",
       "       ['disgust'],\n",
       "       ['surprised'],\n",
       "       ['disgust'],\n",
       "       ['fearful'],\n",
       "       ['sad'],\n",
       "       ['sad'],\n",
       "       ['fearful'],\n",
       "       ['surprised'],\n",
       "       ['sad'],\n",
       "       ['sad'],\n",
       "       ['surprised'],\n",
       "       ['angry'],\n",
       "       ['disgust'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['neutral'],\n",
       "       ['disgust'],\n",
       "       ['neutral'],\n",
       "       ['fearful'],\n",
       "       ['neutral'],\n",
       "       ['neutral'],\n",
       "       ['sad'],\n",
       "       ['angry'],\n",
       "       ['disgust'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['sad'],\n",
       "       ['disgust'],\n",
       "       ['fearful'],\n",
       "       ['disgust'],\n",
       "       ['happy'],\n",
       "       ['disgust'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['angry'],\n",
       "       ['disgust'],\n",
       "       ['sad'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['happy'],\n",
       "       ['fearful'],\n",
       "       ['neutral'],\n",
       "       ['fearful'],\n",
       "       ['angry'],\n",
       "       ['happy'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['disgust'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['neutral'],\n",
       "       ['sad'],\n",
       "       ['surprised'],\n",
       "       ['surprised'],\n",
       "       ['fearful'],\n",
       "       ['sad'],\n",
       "       ['disgust'],\n",
       "       ['fearful'],\n",
       "       ['disgust'],\n",
       "       ['fearful'],\n",
       "       ['sad'],\n",
       "       ['disgust'],\n",
       "       ['disgust'],\n",
       "       ['sad'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['disgust'],\n",
       "       ['sad'],\n",
       "       ['disgust'],\n",
       "       ['surprised'],\n",
       "       ['disgust'],\n",
       "       ['happy'],\n",
       "       ['fearful'],\n",
       "       ['neutral'],\n",
       "       ['angry'],\n",
       "       ['angry'],\n",
       "       ['neutral'],\n",
       "       ['happy'],\n",
       "       ['fearful'],\n",
       "       ['neutral'],\n",
       "       ['surprised'],\n",
       "       ['disgust'],\n",
       "       ['neutral'],\n",
       "       ['angry'],\n",
       "       ['happy'],\n",
       "       ['neutral'],\n",
       "       ['angry'],\n",
       "       ['surprised'],\n",
       "       ['happy'],\n",
       "       ['neutral'],\n",
       "       ['neutral'],\n",
       "       ['neutral'],\n",
       "       ['disgust'],\n",
       "       ['disgust'],\n",
       "       ['angry'],\n",
       "       ['sad'],\n",
       "       ['angry'],\n",
       "       ['disgust'],\n",
       "       ['neutral'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['neutral'],\n",
       "       ['fearful'],\n",
       "       ['angry'],\n",
       "       ['sad'],\n",
       "       ['disgust'],\n",
       "       ['fearful'],\n",
       "       ['happy'],\n",
       "       ['sad'],\n",
       "       ['sad'],\n",
       "       ['angry'],\n",
       "       ['surprised'],\n",
       "       ['fearful'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['disgust'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['neutral'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['disgust'],\n",
       "       ['neutral'],\n",
       "       ['angry'],\n",
       "       ['sad'],\n",
       "       ['surprised'],\n",
       "       ['sad'],\n",
       "       ['angry'],\n",
       "       ['angry'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['disgust'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['disgust'],\n",
       "       ['happy'],\n",
       "       ['fearful'],\n",
       "       ['disgust'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['sad'],\n",
       "       ['disgust'],\n",
       "       ['surprised'],\n",
       "       ['fearful'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['sad'],\n",
       "       ['neutral'],\n",
       "       ['fearful'],\n",
       "       ['neutral'],\n",
       "       ['fearful'],\n",
       "       ['happy'],\n",
       "       ['angry'],\n",
       "       ['disgust'],\n",
       "       ['surprised'],\n",
       "       ['happy'],\n",
       "       ['neutral'],\n",
       "       ['fearful'],\n",
       "       ['sad'],\n",
       "       ['fearful'],\n",
       "       ['disgust'],\n",
       "       ['angry'],\n",
       "       ['happy'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['disgust'],\n",
       "       ['sad'],\n",
       "       ['surprised'],\n",
       "       ['happy'],\n",
       "       ['surprised'],\n",
       "       ['disgust'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['neutral'],\n",
       "       ['sad'],\n",
       "       ['angry'],\n",
       "       ['neutral'],\n",
       "       ['happy'],\n",
       "       ['surprised'],\n",
       "       ['happy'],\n",
       "       ['disgust'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['angry'],\n",
       "       ['disgust'],\n",
       "       ['angry'],\n",
       "       ['neutral'],\n",
       "       ['angry'],\n",
       "       ['happy'],\n",
       "       ['fearful'],\n",
       "       ['happy'],\n",
       "       ['angry'],\n",
       "       ['angry'],\n",
       "       ['happy'],\n",
       "       ['sad'],\n",
       "       ['surprised'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['angry'],\n",
       "       ['happy'],\n",
       "       ['sad'],\n",
       "       ['disgust'],\n",
       "       ['neutral'],\n",
       "       ['neutral'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['surprised'],\n",
       "       ['sad'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['happy'],\n",
       "       ['fearful'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['happy'],\n",
       "       ['fearful'],\n",
       "       ['neutral'],\n",
       "       ['surprised'],\n",
       "       ['sad'],\n",
       "       ['disgust'],\n",
       "       ['surprised'],\n",
       "       ['sad'],\n",
       "       ['angry'],\n",
       "       ['sad'],\n",
       "       ['angry'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['angry'],\n",
       "       ['neutral'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['disgust'],\n",
       "       ['angry'],\n",
       "       ['angry'],\n",
       "       ['angry'],\n",
       "       ['happy'],\n",
       "       ['angry'],\n",
       "       ['surprised'],\n",
       "       ['fearful'],\n",
       "       ['fearful'],\n",
       "       ['surprised'],\n",
       "       ['sad'],\n",
       "       ['happy'],\n",
       "       ['happy'],\n",
       "       ['neutral'],\n",
       "       ['disgust'],\n",
       "       ['disgust'],\n",
       "       ['disgust'],\n",
       "       ['happy'],\n",
       "       ['neutral'],\n",
       "       ['neutral'],\n",
       "       ['neutral'],\n",
       "       ['surprised'],\n",
       "       ['fearful'],\n",
       "       ['neutral'],\n",
       "       ['happy'],\n",
       "       ['angry'],\n",
       "       ['fearful'],\n",
       "       ['disgust'],\n",
       "       ['surprised']], dtype=object)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\users\\kalelt'has\\appdata\\local\\programs\\python\\python38\\lib\\site-packages\\sklearn\\utils\\validation.py:73: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  return f(**kwargs)\n"
     ]
    }
   ],
   "source": [
    "#Label encoding the labels\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "\n",
    "lb = LabelEncoder()\n",
    "\n",
    "y_train = to_categorical(lb.fit_transform(y_train))\n",
    "y_test =  to_categorical(lb.fit_transform(y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_4 (Conv2D)            (None, 216, 12, 20)       520       \n",
      "_________________________________________________________________\n",
      "activation_8 (Activation)    (None, 216, 12, 20)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_4 (MaxPooling2 (None, 108, 6, 20)        0         \n",
      "_________________________________________________________________\n",
      "dropout_6 (Dropout)          (None, 108, 6, 20)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_5 (Conv2D)            (None, 108, 6, 10)        5010      \n",
      "_________________________________________________________________\n",
      "activation_9 (Activation)    (None, 108, 6, 10)        0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_5 (MaxPooling2 (None, 54, 3, 10)         0         \n",
      "_________________________________________________________________\n",
      "dropout_7 (Dropout)          (None, 54, 3, 10)         0         \n",
      "_________________________________________________________________\n",
      "flatten_2 (Flatten)          (None, 1620)              0         \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 1000)              1621000   \n",
      "_________________________________________________________________\n",
      "activation_10 (Activation)   (None, 1000)              0         \n",
      "_________________________________________________________________\n",
      "dropout_8 (Dropout)          (None, 1000)              0         \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 7)                 7007      \n",
      "_________________________________________________________________\n",
      "activation_11 (Activation)   (None, 7)                 0         \n",
      "=================================================================\n",
      "Total params: 1,633,537\n",
      "Trainable params: 1,633,537\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "#Initializing the Model\n",
    "model = Sequential()\n",
    "\n",
    "model.add(Conv2D(20, kernel_size=5, padding=\"same\", input_shape=(216, 12 ,1)))\n",
    "model.add(Activation('relu'))\n",
    "model.add(MaxPooling2D(pool_size=(2,2), strides=(2,2)))\n",
    "model.add(Dropout(0.1))\n",
    "\n",
    "model.add(Conv2D(10, kernel_size=5, padding=\"same\"))\n",
    "model.add(Activation('relu'))\n",
    "model.add(MaxPooling2D(pool_size=(2,2), strides=(2,2)))\n",
    "model.add(Dropout(0.1))\n",
    "\n",
    "model.add(Flatten())\n",
    "\n",
    "model.add(Dense(1000))\n",
    "model.add(Activation('relu'))\n",
    "model.add(Dropout(0.1))\n",
    "\n",
    "model.add(Dense(7))\n",
    "model.add(Activation('softmax'))\n",
    "\n",
    "model.summary()\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [],
   "source": [
    "model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/105\n",
      "102/102 [==============================] - 30s 292ms/step - loss: 0.4092 - accuracy: 0.1884 - val_loss: 0.3955 - val_accuracy: 0.2220\n",
      "Epoch 2/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.3804 - accuracy: 0.2709 - val_loss: 0.3917 - val_accuracy: 0.2799\n",
      "Epoch 3/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.3692 - accuracy: 0.3013 - val_loss: 0.3650 - val_accuracy: 0.2934\n",
      "Epoch 4/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.3498 - accuracy: 0.3626 - val_loss: 0.3415 - val_accuracy: 0.3842\n",
      "Epoch 5/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.3367 - accuracy: 0.3876 - val_loss: 0.3650 - val_accuracy: 0.3263\n",
      "Epoch 6/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.3301 - accuracy: 0.4048 - val_loss: 0.3448 - val_accuracy: 0.3436\n",
      "Epoch 7/105\n",
      "102/102 [==============================] - 29s 286ms/step - loss: 0.3245 - accuracy: 0.4176 - val_loss: 0.3260 - val_accuracy: 0.3958\n",
      "Epoch 8/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.3149 - accuracy: 0.4485 - val_loss: 0.3253 - val_accuracy: 0.4015\n",
      "Epoch 9/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.3078 - accuracy: 0.4632 - val_loss: 0.3120 - val_accuracy: 0.4537\n",
      "Epoch 10/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.3003 - accuracy: 0.4769 - val_loss: 0.3352 - val_accuracy: 0.4112\n",
      "Epoch 11/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.2963 - accuracy: 0.4814 - val_loss: 0.3033 - val_accuracy: 0.4653\n",
      "Epoch 12/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.2914 - accuracy: 0.4858 - val_loss: 0.3128 - val_accuracy: 0.4556\n",
      "Epoch 13/105\n",
      "102/102 [==============================] - 29s 287ms/step - loss: 0.2852 - accuracy: 0.5083 - val_loss: 0.3094 - val_accuracy: 0.4479\n",
      "Epoch 14/105\n",
      "102/102 [==============================] - 29s 286ms/step - loss: 0.2752 - accuracy: 0.5280 - val_loss: 0.3016 - val_accuracy: 0.4595\n",
      "Epoch 15/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.2703 - accuracy: 0.5496 - val_loss: 0.3051 - val_accuracy: 0.4749\n",
      "Epoch 16/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.2678 - accuracy: 0.5324 - val_loss: 0.3030 - val_accuracy: 0.4749\n",
      "Epoch 17/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.2672 - accuracy: 0.5363 - val_loss: 0.2902 - val_accuracy: 0.5019\n",
      "Epoch 18/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.2561 - accuracy: 0.5726 - val_loss: 0.2932 - val_accuracy: 0.4961\n",
      "Epoch 19/105\n",
      "102/102 [==============================] - 29s 286ms/step - loss: 0.2517 - accuracy: 0.5800 - val_loss: 0.2926 - val_accuracy: 0.5077\n",
      "Epoch 20/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.2485 - accuracy: 0.5726 - val_loss: 0.2830 - val_accuracy: 0.5154\n",
      "Epoch 21/105\n",
      "102/102 [==============================] - 31s 299ms/step - loss: 0.2412 - accuracy: 0.5981 - val_loss: 0.2822 - val_accuracy: 0.5039\n",
      "Epoch 22/105\n",
      "102/102 [==============================] - 30s 295ms/step - loss: 0.2377 - accuracy: 0.6055 - val_loss: 0.2942 - val_accuracy: 0.5058\n",
      "Epoch 23/105\n",
      "102/102 [==============================] - 29s 289ms/step - loss: 0.2290 - accuracy: 0.6050 - val_loss: 0.3202 - val_accuracy: 0.4208\n",
      "Epoch 24/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.2299 - accuracy: 0.6173 - val_loss: 0.2897 - val_accuracy: 0.5328\n",
      "Epoch 25/105\n",
      "102/102 [==============================] - 29s 284ms/step - loss: 0.2187 - accuracy: 0.6379 - val_loss: 0.3012 - val_accuracy: 0.4961\n",
      "Epoch 26/105\n",
      "102/102 [==============================] - 29s 286ms/step - loss: 0.2195 - accuracy: 0.6364 - val_loss: 0.2829 - val_accuracy: 0.5328\n",
      "Epoch 27/105\n",
      "102/102 [==============================] - 29s 287ms/step - loss: 0.2109 - accuracy: 0.6590 - val_loss: 0.2729 - val_accuracy: 0.5541\n",
      "Epoch 28/105\n",
      "102/102 [==============================] - 29s 286ms/step - loss: 0.2002 - accuracy: 0.6742 - val_loss: 0.3028 - val_accuracy: 0.5347\n",
      "Epoch 29/105\n",
      "102/102 [==============================] - 29s 287ms/step - loss: 0.1971 - accuracy: 0.6791 - val_loss: 0.2910 - val_accuracy: 0.5077\n",
      "Epoch 30/105\n",
      "102/102 [==============================] - 29s 287ms/step - loss: 0.1881 - accuracy: 0.6919 - val_loss: 0.2878 - val_accuracy: 0.5154\n",
      "Epoch 31/105\n",
      "102/102 [==============================] - 29s 287ms/step - loss: 0.1882 - accuracy: 0.7002 - val_loss: 0.2742 - val_accuracy: 0.5444\n",
      "Epoch 32/105\n",
      "102/102 [==============================] - 29s 287ms/step - loss: 0.1794 - accuracy: 0.7095 - val_loss: 0.2948 - val_accuracy: 0.5097\n",
      "Epoch 33/105\n",
      "102/102 [==============================] - 30s 294ms/step - loss: 0.1732 - accuracy: 0.7355 - val_loss: 0.2930 - val_accuracy: 0.5367\n",
      "Epoch 34/105\n",
      "102/102 [==============================] - 29s 288ms/step - loss: 0.1733 - accuracy: 0.7301 - val_loss: 0.2884 - val_accuracy: 0.5541\n",
      "Epoch 35/105\n",
      "102/102 [==============================] - 29s 282ms/step - loss: 0.1653 - accuracy: 0.7370 - val_loss: 0.3029 - val_accuracy: 0.5309\n",
      "Epoch 36/105\n",
      "102/102 [==============================] - 29s 284ms/step - loss: 0.1613 - accuracy: 0.7458 - val_loss: 0.2958 - val_accuracy: 0.5618\n",
      "Epoch 37/105\n",
      "102/102 [==============================] - 31s 303ms/step - loss: 0.1590 - accuracy: 0.7532 - val_loss: 0.2836 - val_accuracy: 0.5598\n",
      "Epoch 38/105\n",
      "102/102 [==============================] - 32s 318ms/step - loss: 0.1501 - accuracy: 0.7679 - val_loss: 0.2925 - val_accuracy: 0.5637\n",
      "Epoch 39/105\n",
      "102/102 [==============================] - 33s 319ms/step - loss: 0.1470 - accuracy: 0.7689 - val_loss: 0.3206 - val_accuracy: 0.5290\n",
      "Epoch 40/105\n",
      "102/102 [==============================] - 32s 316ms/step - loss: 0.1474 - accuracy: 0.7669 - val_loss: 0.2831 - val_accuracy: 0.5714\n",
      "Epoch 41/105\n",
      "102/102 [==============================] - 35s 341ms/step - loss: 0.1335 - accuracy: 0.7920 - val_loss: 0.3001 - val_accuracy: 0.5676\n",
      "Epoch 42/105\n",
      "102/102 [==============================] - 36s 350ms/step - loss: 0.1348 - accuracy: 0.7929 - val_loss: 0.2951 - val_accuracy: 0.5907\n",
      "Epoch 43/105\n",
      "102/102 [==============================] - 39s 384ms/step - loss: 0.1288 - accuracy: 0.8116 - val_loss: 0.3112 - val_accuracy: 0.5695\n",
      "Epoch 44/105\n",
      "102/102 [==============================] - 38s 368ms/step - loss: 0.1133 - accuracy: 0.8292 - val_loss: 0.3278 - val_accuracy: 0.5618\n",
      "Epoch 50/105\n",
      "102/102 [==============================] - 35s 347ms/step - loss: 0.1050 - accuracy: 0.8435 - val_loss: 0.3130 - val_accuracy: 0.5541\n",
      "Epoch 51/105\n",
      "102/102 [==============================] - 36s 355ms/step - loss: 0.1060 - accuracy: 0.8538 - val_loss: 0.3158 - val_accuracy: 0.5676\n",
      "Epoch 52/105\n",
      "102/102 [==============================] - 36s 349ms/step - loss: 0.0954 - accuracy: 0.8567 - val_loss: 0.3238 - val_accuracy: 0.5888\n",
      "Epoch 53/105\n",
      "102/102 [==============================] - 36s 348ms/step - loss: 0.0948 - accuracy: 0.8592 - val_loss: 0.3590 - val_accuracy: 0.5695\n",
      "Epoch 54/105\n",
      "102/102 [==============================] - 36s 354ms/step - loss: 0.0928 - accuracy: 0.8567 - val_loss: 0.3314 - val_accuracy: 0.5695\n",
      "Epoch 55/105\n",
      "102/102 [==============================] - 36s 357ms/step - loss: 0.0899 - accuracy: 0.8739 - val_loss: 0.3329 - val_accuracy: 0.5618\n",
      "Epoch 56/105\n",
      "102/102 [==============================] - 36s 353ms/step - loss: 0.0928 - accuracy: 0.8606 - val_loss: 0.3251 - val_accuracy: 0.5753\n",
      "Epoch 57/105\n",
      "102/102 [==============================] - 36s 355ms/step - loss: 0.0855 - accuracy: 0.8783 - val_loss: 0.3316 - val_accuracy: 0.5772\n",
      "Epoch 58/105\n",
      "102/102 [==============================] - 36s 350ms/step - loss: 0.0877 - accuracy: 0.8729 - val_loss: 0.3582 - val_accuracy: 0.5521\n",
      "Epoch 59/105\n",
      "102/102 [==============================] - 36s 353ms/step - loss: 0.0789 - accuracy: 0.8847 - val_loss: 0.3477 - val_accuracy: 0.5656\n",
      "Epoch 60/105\n",
      "102/102 [==============================] - 36s 354ms/step - loss: 0.0706 - accuracy: 0.9014 - val_loss: 0.3676 - val_accuracy: 0.5656\n",
      "Epoch 61/105\n",
      "102/102 [==============================] - 34s 336ms/step - loss: 0.0755 - accuracy: 0.8930 - val_loss: 0.3698 - val_accuracy: 0.5656\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 62/105\n",
      "102/102 [==============================] - 35s 342ms/step - loss: 0.0746 - accuracy: 0.8989 - val_loss: 0.3479 - val_accuracy: 0.5985\n",
      "Epoch 63/105\n",
      "102/102 [==============================] - 34s 336ms/step - loss: 0.0662 - accuracy: 0.9092 - val_loss: 0.3658 - val_accuracy: 0.6120\n",
      "Epoch 64/105\n",
      "102/102 [==============================] - 35s 340ms/step - loss: 0.0633 - accuracy: 0.9082 - val_loss: 0.3653 - val_accuracy: 0.5714\n",
      "Epoch 65/105\n",
      "102/102 [==============================] - 35s 340ms/step - loss: 0.0773 - accuracy: 0.8921 - val_loss: 0.3576 - val_accuracy: 0.5714\n",
      "Epoch 66/105\n",
      "102/102 [==============================] - 36s 355ms/step - loss: 0.0699 - accuracy: 0.9068 - val_loss: 0.3944 - val_accuracy: 0.5618\n",
      "Epoch 67/105\n",
      "102/102 [==============================] - 37s 362ms/step - loss: 0.0606 - accuracy: 0.9146 - val_loss: 0.3767 - val_accuracy: 0.5946\n",
      "Epoch 68/105\n",
      "102/102 [==============================] - 38s 374ms/step - loss: 0.0578 - accuracy: 0.9161 - val_loss: 0.3904 - val_accuracy: 0.6023\n",
      "Epoch 69/105\n",
      "102/102 [==============================] - 38s 373ms/step - loss: 0.0576 - accuracy: 0.9235 - val_loss: 0.3857 - val_accuracy: 0.5772\n",
      "Epoch 70/105\n",
      "102/102 [==============================] - 37s 360ms/step - loss: 0.0548 - accuracy: 0.9230 - val_loss: 0.3821 - val_accuracy: 0.6100\n",
      "Epoch 73/105\n",
      "102/102 [==============================] - 36s 353ms/step - loss: 0.0568 - accuracy: 0.9205 - val_loss: 0.4282 - val_accuracy: 0.5753\n",
      "Epoch 74/105\n",
      "102/102 [==============================] - 36s 352ms/step - loss: 0.0478 - accuracy: 0.9347 - val_loss: 0.4056 - val_accuracy: 0.5907\n",
      "Epoch 75/105\n",
      "102/102 [==============================] - 34s 329ms/step - loss: 0.0683 - accuracy: 0.9038 - val_loss: 0.4433 - val_accuracy: 0.5714\n",
      "Epoch 76/105\n",
      "102/102 [==============================] - 35s 343ms/step - loss: 0.0561 - accuracy: 0.9181 - val_loss: 0.4328 - val_accuracy: 0.5849\n",
      "Epoch 77/105\n",
      "102/102 [==============================] - 34s 334ms/step - loss: 0.0481 - accuracy: 0.9338 - val_loss: 0.3921 - val_accuracy: 0.6023\n",
      "Epoch 78/105\n",
      "102/102 [==============================] - 30s 298ms/step - loss: 0.0455 - accuracy: 0.9352 - val_loss: 0.3865 - val_accuracy: 0.6197\n",
      "Epoch 79/105\n",
      "102/102 [==============================] - 29s 283ms/step - loss: 0.0442 - accuracy: 0.9470 - val_loss: 0.4248 - val_accuracy: 0.6081\n",
      "Epoch 80/105\n",
      "102/102 [==============================] - 29s 285ms/step - loss: 0.0412 - accuracy: 0.9465 - val_loss: 0.4119 - val_accuracy: 0.6100\n",
      "Epoch 81/105\n",
      "102/102 [==============================] - 27s 269ms/step - loss: 0.0409 - accuracy: 0.9406 - val_loss: 0.4447 - val_accuracy: 0.6100\n",
      "Epoch 82/105\n",
      "102/102 [==============================] - 28s 270ms/step - loss: 0.0439 - accuracy: 0.9426 - val_loss: 0.4387 - val_accuracy: 0.5869\n",
      "Epoch 83/105\n",
      "102/102 [==============================] - 27s 269ms/step - loss: 0.0513 - accuracy: 0.9274 - val_loss: 0.4024 - val_accuracy: 0.6178\n",
      "Epoch 84/105\n",
      "102/102 [==============================] - 28s 272ms/step - loss: 0.0387 - accuracy: 0.9470 - val_loss: 0.4050 - val_accuracy: 0.6274\n",
      "Epoch 85/105\n",
      "102/102 [==============================] - 27s 266ms/step - loss: 0.0385 - accuracy: 0.9544 - val_loss: 0.4177 - val_accuracy: 0.6023\n",
      "Epoch 86/105\n",
      "102/102 [==============================] - 28s 272ms/step - loss: 0.0383 - accuracy: 0.9460 - val_loss: 0.4953 - val_accuracy: 0.5830\n",
      "Epoch 87/105\n",
      "102/102 [==============================] - 28s 272ms/step - loss: 0.0365 - accuracy: 0.9514 - val_loss: 0.4307 - val_accuracy: 0.5927\n",
      "Epoch 88/105\n",
      "102/102 [==============================] - 27s 265ms/step - loss: 0.0365 - accuracy: 0.9529 - val_loss: 0.4189 - val_accuracy: 0.5985\n",
      "Epoch 89/105\n",
      "102/102 [==============================] - 27s 266ms/step - loss: 0.0402 - accuracy: 0.9450 - val_loss: 0.4114 - val_accuracy: 0.6197\n",
      "Epoch 90/105\n",
      "102/102 [==============================] - 28s 276ms/step - loss: 0.0362 - accuracy: 0.9519 - val_loss: 0.4443 - val_accuracy: 0.6197\n",
      "Epoch 91/105\n",
      "102/102 [==============================] - 28s 276ms/step - loss: 0.0404 - accuracy: 0.9411 - val_loss: 0.4090 - val_accuracy: 0.6313\n",
      "Epoch 92/105\n",
      "102/102 [==============================] - 27s 267ms/step - loss: 0.0313 - accuracy: 0.9549 - val_loss: 0.4121 - val_accuracy: 0.6293\n",
      "Epoch 93/105\n",
      "102/102 [==============================] - 27s 265ms/step - loss: 0.0302 - accuracy: 0.9588 - val_loss: 0.4268 - val_accuracy: 0.6255\n",
      "Epoch 94/105\n",
      "102/102 [==============================] - 28s 276ms/step - loss: 0.0344 - accuracy: 0.9534 - val_loss: 0.4326 - val_accuracy: 0.5907\n",
      "Epoch 95/105\n",
      "102/102 [==============================] - 29s 280ms/step - loss: 0.0432 - accuracy: 0.9406 - val_loss: 0.3821 - val_accuracy: 0.6313\n",
      "Epoch 96/105\n",
      "102/102 [==============================] - 29s 282ms/step - loss: 0.0275 - accuracy: 0.9696 - val_loss: 0.4310 - val_accuracy: 0.6178\n",
      "Epoch 97/105\n",
      "102/102 [==============================] - 27s 269ms/step - loss: 0.0354 - accuracy: 0.9563 - val_loss: 0.4761 - val_accuracy: 0.6081\n",
      "Epoch 98/105\n",
      "102/102 [==============================] - 27s 268ms/step - loss: 0.0332 - accuracy: 0.9568 - val_loss: 0.4564 - val_accuracy: 0.6081\n",
      "Epoch 99/105\n",
      "102/102 [==============================] - 29s 283ms/step - loss: 0.0344 - accuracy: 0.9539 - val_loss: 0.4387 - val_accuracy: 0.6120\n",
      "Epoch 100/105\n",
      "102/102 [==============================] - 28s 272ms/step - loss: 0.0326 - accuracy: 0.9568 - val_loss: 0.4261 - val_accuracy: 0.6081\n",
      "Epoch 101/105\n",
      "102/102 [==============================] - 27s 265ms/step - loss: 0.0276 - accuracy: 0.9676 - val_loss: 0.4650 - val_accuracy: 0.5985\n",
      "Epoch 102/105\n",
      "102/102 [==============================] - 27s 267ms/step - loss: 0.0254 - accuracy: 0.9671 - val_loss: 0.4560 - val_accuracy: 0.5985\n",
      "Epoch 103/105\n",
      "102/102 [==============================] - 27s 265ms/step - loss: 0.0328 - accuracy: 0.9598 - val_loss: 0.5321 - val_accuracy: 0.5695\n",
      "Epoch 104/105\n",
      "102/102 [==============================] - 27s 266ms/step - loss: 0.0234 - accuracy: 0.9701 - val_loss: 0.5381 - val_accuracy: 0.5734\n",
      "Epoch 105/105\n",
      "102/102 [==============================] - 27s 266ms/step - loss: 0.0342 - accuracy: 0.9524 - val_loss: 0.4899 - val_accuracy: 0.6023\n"
     ]
    }
   ],
   "source": [
    "cnnhistory = model.fit(X_train_cnn, y_train, batch_size=20, epochs=105, verbose=1, validation_data=(X_test_cnn, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {
    "pycharm": {
     "is_executing": true
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(cnnhistory.history['loss'])\n",
    "plt.plot(cnnhistory.history['val_loss'])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "17/17 [==============================] - 2s 108ms/step - loss: 0.4899 - accuracy: 0.6023\n"
     ]
    }
   ],
   "source": [
    "scores = model.evaluate(X_test_cnn, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accuracy: 60.23%\n"
     ]
    }
   ],
   "source": [
    "print(\"%s: %.2f%%\" % (model.metrics_names[1], scores[1]*100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {
    "pycharm": {
     "is_executing": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saved trained model at C:\\Users\\Kalelt'has\\Desktop\\Stage\\ASP\\saved_models\\Emotion_Voice_Detection_Model.h5 \n"
     ]
    }
   ],
   "source": [
    "model_name = 'Emotion_Voice_Detection_Model.h5'\n",
    "save_dir = os.path.join(os.getcwd(), 'saved_models')\n",
    "# Save model and weights\n",
    "if not os.path.isdir(save_dir):\n",
    "    os.makedirs(save_dir)\n",
    "model_path = os.path.join(save_dir, model_name)\n",
    "model.save(model_path)\n",
    "print('Saved trained model at %s ' % model_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "model_json = model.to_json()\n",
    "with open(\"model.json\", \"w\") as json_file:\n",
    "    json_file.write(model_json)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loaded model from disk\n"
     ]
    }
   ],
   "source": [
    "# loading json and creating model\n",
    "from keras.models import model_from_json\n",
    "json_file = open('model.json', 'r')\n",
    "loaded_model_json = json_file.read()\n",
    "json_file.close()\n",
    "loaded_model = model_from_json(loaded_model_json)\n",
    "# load weights into new model\n",
    "loaded_model.load_weights(\"saved_models/Emotion_Voice_Detection_Model.h5\")\n",
    "print(\"Loaded model from disk\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accuracy: 92.62%\n"
     ]
    }
   ],
   "source": [
    "# evaluate loaded model on test data\n",
    "loaded_model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])\n",
    "score = loaded_model.evaluate(X_test_cnn, y_test, verbose=0)\n",
    "print(\"%s: %.2f%%\" % (loaded_model.metrics_names[1], score[1]*100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "17/17 [==============================] - 1s 48ms/step\n"
     ]
    },
    {
     "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>predictedvalues</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>actualvalues</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>angry</th>\n",
       "      <td>92</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>disgust</th>\n",
       "      <td>63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fearful</th>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>happy</th>\n",
       "      <td>86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>neutral</th>\n",
       "      <td>62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sad</th>\n",
       "      <td>82</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>surprised</th>\n",
       "      <td>53</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              predictedvalues\n",
       "actualvalues                 \n",
       "angry                      92\n",
       "disgust                    63\n",
       "fearful                    95\n",
       "happy                      86\n",
       "neutral                    62\n",
       "sad                        82\n",
       "surprised                  53"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preds = loaded_model.predict(X_test_cnn, \n",
    "                         batch_size=32, \n",
    "                         verbose=1)\n",
    "preds1=preds.argmax(axis=1)\n",
    "abc = preds1.astype(int).flatten()\n",
    "predictions = (lb.inverse_transform((abc)))\n",
    "preddf = pd.DataFrame({'predictedvalues': predictions})\n",
    "actual=y_test.argmax(axis=1)\n",
    "abc123 = actual.astype(int).flatten()\n",
    "actualvalues = (lb.inverse_transform((abc123)))\n",
    "actualdf = pd.DataFrame({'actualvalues': actualvalues})\n",
    "finaldf = actualdf.join(preddf)\n",
    "finaldf.groupby('actualvalues').count()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "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>actualvalues</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>predictedvalues</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>angry</th>\n",
       "      <td>88</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>disgust</th>\n",
       "      <td>61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fearful</th>\n",
       "      <td>94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>happy</th>\n",
       "      <td>86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>neutral</th>\n",
       "      <td>64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sad</th>\n",
       "      <td>86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>surprised</th>\n",
       "      <td>54</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 actualvalues\n",
       "predictedvalues              \n",
       "angry                      88\n",
       "disgust                    61\n",
       "fearful                    94\n",
       "happy                      86\n",
       "neutral                    64\n",
       "sad                        86\n",
       "surprised                  54"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "finaldf.groupby('predictedvalues').count()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "data, sampling_rate = librosa.load('C:\\\\Users\\Kalelt\\'has\\Desktop\\Stage\\ASP\\saved_voice\\\\output10.wav')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PolyCollection at 0x27fb5b4be50>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import os\n",
    "import pandas as pd\n",
    "import librosa\n",
    "import glob \n",
    "\n",
    "plt.figure(figsize=(15, 5))\n",
    "librosa.display.waveplot(data, sr=sampling_rate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Prepare the dict for the recorded sample\n",
    "my_dict = {}\n",
    "i=0\n",
    "for i in range(216):\n",
    "    my_dict[i] = np.zeros(12)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "X, sample_rate = librosa.load('C:\\\\Users\\Kalelt\\'has\\Desktop\\Stage\\ASP\\saved_voice\\\\record_surprised.wav', res_type='kaiser_fast',duration=2.5,sr=22050*2,offset=0.5)\n",
    "sample_rate = np.array(sample_rate)\n",
    "mfccs = librosa.feature.mfcc(y=X,sr=sample_rate,n_mfcc=12).T\n",
    "for i in range(len(mfccs)):\n",
    "    my_dict[i] = mfccs[i]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_record = pd.DataFrame.from_dict(my_dict, orient='index')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(216, 12)"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_record = np.array(df_record)\n",
    "X_record.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_record = X_record.reshape(1,216,12,1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "livepreds = loaded_model.predict(X_record, \n",
    "                         batch_size=32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 0., 1., 0., 0., 0., 0.]], dtype=float32)"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "livepreds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['fearful'], dtype=object)"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "livepreds1=livepreds.argmax(axis=1)\n",
    "liveabc = livepreds1.astype(int).flatten()\n",
    "livepredictions = (lb.inverse_transform((liveabc)))\n",
    "livepredictions\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  },
  "pycharm": {
   "stem_cell": {
    "cell_type": "raw",
    "source": [],
    "metadata": {
     "collapsed": false
    }
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}