panda_read_CSV.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd \n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt \n",
    "import matplotlib.image as mpimg\n",
    "\n",
    "from sklearn import svm, metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Read csv file\n",
    "data = pd.read_csv('fer2013/fer2013.csv')\n",
    "\n",
    "#Number of samples\n",
    "n_samples = len(data)\n",
    "n_samples_train = 28709\n",
    "n_samples_test = 3589\n",
    "n_samples_validation = 3589\n",
    "\n",
    "#Pixel width and height\n",
    "w = 48\n",
    "h = 48"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#Separating labels and features respectively\n",
    "y = data['emotion']\n",
    "X = np.zeros((n_samples, w, h))\n",
    "for i in range(n_samples):\n",
    "    X[i] = np.fromstring(data['pixels'][i], dtype=int, sep=' ').reshape(w, h)\n",
    "#Training set \n",
    "X_train = X[:n_samples_train].reshape(n_samples_train, -1)\n",
    "y_train = y[:n_samples_train]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "anger_train = data.loc[(data['Usage'] == 'PrivateTest') &  (data['emotion'] == 6 )]\n",
    "anger_reset = anger_train.reset_index()\n",
    "nt_samples = len(anger_reset)\n",
    "\n",
    "y3 = np.zeros((nt_samples, w, h))\n",
    "for j in range(nt_samples):\n",
    "    y3[j] = np.fromstring(anger_reset['pixels'][j], dtype=int, sep=' ').reshape(w,h)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from PIL import Image\n",
    "for i in range(nt_samples):\n",
    "    im = Image.fromarray(y3[i])\n",
    "    if im.mode != 'RGB':\n",
    "        im = im.convert('RGB')\n",
    "    im.save(\"test/neutral/neutral\" + str(i) + \".png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEF/4jpNDfOE7Tg7xhe84OeT/AcPOA3JWNPN9AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f9d90492828>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "import matplotlib.pyplot as plt\n",
    "plt.show()\n",
    "plt.imshow(X[0]) #Needs to be in row,col order\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from PIL import Image\n",
    "for i in range(1,10):\n",
    "    im = Image.fromarray(X[i])\n",
    "    if im.mode != 'RGB':\n",
    "        im = im.convert('RGB')\n",
    "    im.save(\"test\" + str(i) + \".png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}