let's_explore_fer2013.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv('data/fer2013.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Description\n",
    "* emotion label (0=Angry, 1=Disgust, 2=Fear, 3=Happy, 4=Sad, 5=Surprise, 6=Neutral)\n",
    "* 48 X 48 pixels \n",
    "* The training set consists of 28,709 examples. <br>The public test set used for the leaderboard consists of 3,589 examples.<br>The final test set, which was used to determine the winner of the competition, consists of another 3,589 examples."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>emotion</th>\n",
       "      <th>pixels</th>\n",
       "      <th>Usage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>70 80 82 72 58 58 60 63 54 58 60 48 89 115 121...</td>\n",
       "      <td>Training</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>151 150 147 155 148 133 111 140 170 174 182 15...</td>\n",
       "      <td>Training</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>231 212 156 164 174 138 161 173 182 200 106 38...</td>\n",
       "      <td>Training</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>24 32 36 30 32 23 19 20 30 41 21 22 32 34 21 1...</td>\n",
       "      <td>Training</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>6</td>\n",
       "      <td>4 0 0 0 0 0 0 0 0 0 0 0 3 15 23 28 48 50 58 84...</td>\n",
       "      <td>Training</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   emotion                                             pixels     Usage\n",
       "0        0  70 80 82 72 58 58 60 63 54 58 60 48 89 115 121...  Training\n",
       "1        0  151 150 147 155 148 133 111 140 170 174 182 15...  Training\n",
       "2        2  231 212 156 164 174 138 161 173 182 200 106 38...  Training\n",
       "3        4  24 32 36 30 32 23 19 20 30 41 21 22 32 34 21 1...  Training\n",
       "4        6  4 0 0 0 0 0 0 0 0 0 0 0 3 15 23 28 48 50 58 84...  Training"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>emotion</th>\n",
       "      <th>pixels</th>\n",
       "      <th>Usage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>35882</th>\n",
       "      <td>6</td>\n",
       "      <td>50 36 17 22 23 29 33 39 34 37 37 37 39 43 48 5...</td>\n",
       "      <td>PrivateTest</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35883</th>\n",
       "      <td>3</td>\n",
       "      <td>178 174 172 173 181 188 191 194 196 199 200 20...</td>\n",
       "      <td>PrivateTest</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35884</th>\n",
       "      <td>0</td>\n",
       "      <td>17 17 16 23 28 22 19 17 25 26 20 24 31 19 27 9...</td>\n",
       "      <td>PrivateTest</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35885</th>\n",
       "      <td>3</td>\n",
       "      <td>30 28 28 29 31 30 42 68 79 81 77 67 67 71 63 6...</td>\n",
       "      <td>PrivateTest</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35886</th>\n",
       "      <td>2</td>\n",
       "      <td>19 13 14 12 13 16 21 33 50 57 71 84 97 108 122...</td>\n",
       "      <td>PrivateTest</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       emotion                                             pixels        Usage\n",
       "35882        6  50 36 17 22 23 29 33 39 34 37 37 37 39 43 48 5...  PrivateTest\n",
       "35883        3  178 174 172 173 181 188 191 194 196 199 200 20...  PrivateTest\n",
       "35884        0  17 17 16 23 28 22 19 17 25 26 20 24 31 19 27 9...  PrivateTest\n",
       "35885        3  30 28 28 29 31 30 42 68 79 81 77 67 67 71 63 6...  PrivateTest\n",
       "35886        2  19 13 14 12 13 16 21 33 50 57 71 84 97 108 122...  PrivateTest"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "35887"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.shape[0] # whole data size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "pixels = data['pixels'].tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "faces = []\n",
    "\n",
    "for pixel_row in pixels:\n",
    "    face = [int(pixel) for pixel in pixel_row.split(' ')]\n",
    "    face = np.asarray(face).reshape(48, 48) # pixel size\n",
    "    faces.append(face.astype('float32'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "faces = np.asarray(faces)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[ 70.,  80.,  82., ...,  52.,  43.,  41.],\n",
       "        [ 65.,  61.,  58., ...,  56.,  52.,  44.],\n",
       "        [ 50.,  43.,  54., ...,  49.,  56.,  47.],\n",
       "        ...,\n",
       "        [ 91.,  65.,  42., ...,  72.,  56.,  43.],\n",
       "        [ 77.,  82.,  79., ..., 105.,  70.,  46.],\n",
       "        [ 77.,  72.,  84., ..., 106., 109.,  82.]],\n",
       "\n",
       "       [[151., 150., 147., ..., 129., 140., 120.],\n",
       "        [151., 149., 149., ..., 122., 141., 137.],\n",
       "        [151., 151., 156., ..., 109., 123., 146.],\n",
       "        ...,\n",
       "        [188., 188., 121., ..., 185., 185., 186.],\n",
       "        [188., 187., 196., ..., 186., 182., 187.],\n",
       "        [186., 184., 185., ..., 193., 183., 184.]],\n",
       "\n",
       "       [[231., 212., 156., ...,  44.,  27.,  16.],\n",
       "        [229., 175., 148., ...,  27.,  35.,  27.],\n",
       "        [214., 156., 157., ...,  28.,  22.,  28.],\n",
       "        ...,\n",
       "        [241., 245., 250., ...,  57., 101., 146.],\n",
       "        [246., 250., 252., ...,  78., 105., 162.],\n",
       "        [250., 251., 250., ...,  88., 110., 152.]],\n",
       "\n",
       "       ...,\n",
       "\n",
       "       [[ 17.,  17.,  16., ...,  83., 114., 245.],\n",
       "        [ 18.,  17.,  16., ..., 104., 136., 253.],\n",
       "        [ 19.,  16.,  17., ..., 128., 152., 255.],\n",
       "        ...,\n",
       "        [  4.,  21.,  46., ..., 186., 180., 187.],\n",
       "        [  5.,  17.,  41., ..., 177., 172., 176.],\n",
       "        [ 20.,  15.,  22., ..., 154., 133., 113.]],\n",
       "\n",
       "       [[ 30.,  28.,  28., ...,  60.,  50.,  44.],\n",
       "        [ 30.,  27.,  28., ...,  64.,  52.,  40.],\n",
       "        [ 31.,  28.,  30., ...,  61.,  54.,  37.],\n",
       "        ...,\n",
       "        [104., 109., 110., ...,  35.,  30.,  30.],\n",
       "        [102., 105., 108., ...,  35.,  31.,  29.],\n",
       "        [ 93.,  96., 100., ...,  35.,  30.,  28.]],\n",
       "\n",
       "       [[ 19.,  13.,  14., ..., 108.,  95.,  86.],\n",
       "        [ 16.,  17.,  15., ..., 105.,  94.,  90.],\n",
       "        [ 10.,   9.,  10., ..., 101.,  93.,  95.],\n",
       "        ...,\n",
       "        [ 18.,  14.,  16., ...,  55.,  64.,  95.],\n",
       "        [ 15.,  15.,  13., ..., 123., 171., 192.],\n",
       "        [ 16.,  14.,  13., ..., 189., 199., 201.]]], dtype=float32)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "faces"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 70.,  80.,  82., ...,  52.,  43.,  41.],\n",
       "       [ 65.,  61.,  58., ...,  56.,  52.,  44.],\n",
       "       [ 50.,  43.,  54., ...,  49.,  56.,  47.],\n",
       "       ...,\n",
       "       [ 91.,  65.,  42., ...,  72.,  56.,  43.],\n",
       "       [ 77.,  82.,  79., ..., 105.,  70.,  46.],\n",
       "       [ 77.,  72.,  84., ..., 106., 109.,  82.]], dtype=float32)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "faces[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(48, 48)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "faces[0].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x112405a90>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(faces[0], cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x11220b9e8>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(faces[701], cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data = data[data['Usage']=='Training']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "28709"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.shape[0] # train data size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3    7215\n",
       "6    4965\n",
       "4    4830\n",
       "2    4097\n",
       "0    3995\n",
       "5    3171\n",
       "1     436\n",
       "Name: emotion, dtype: int64"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data['emotion'].value_counts() # Show ratio by label in train data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1, 0, 0, ..., 0, 0, 0],\n",
       "       [1, 0, 0, ..., 0, 0, 0],\n",
       "       [0, 0, 1, ..., 0, 0, 0],\n",
       "       ...,\n",
       "       [1, 0, 0, ..., 0, 0, 0],\n",
       "       [0, 0, 0, ..., 0, 0, 0],\n",
       "       [0, 0, 1, ..., 0, 0, 0]], dtype=uint8)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "faces = np.asarray(faces)\n",
    "faces = np.expand_dims(faces, -1) \n",
    "\n",
    "emotions = pd.get_dummies(data['emotion']).values\n",
    "emotions"
   ]
  }
 ],
 "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}