irisdataset_with_deeplerning.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# This Python 3 environment comes with many helpful analytics libraries installed\n",
    "# It is defined by the kaggle/python docker image: https://github.com/kaggle/docker-python\n",
    "# For example, here's several helpful packages to load in \n",
    "\n",
    "import numpy as np # linear algebra\n",
    "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n",
    "#import seaborn as sns # visualization\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "# Input data files are available in the \"../input/\" directory.\n",
    "# For example, running this (by clicking run or pressing Shift+Enter) will list the files in the input directory\n",
    "\n",
    "#from subprocess import check_output\n",
    "#print(check_output([\"ls\", \"../input\"]).decode(\"utf8\"))\n",
    "\n",
    "# Any results you write to the current directory are saved as output.\n",
    "#https://www.kaggle.com/ancientaxe/simple-neural-network-from-scratch-in-python/notebook"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Id</th>\n",
       "      <th>SepalLengthCm</th>\n",
       "      <th>SepalWidthCm</th>\n",
       "      <th>PetalLengthCm</th>\n",
       "      <th>PetalWidthCm</th>\n",
       "      <th>Species</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>16</td>\n",
       "      <td>5.7</td>\n",
       "      <td>4.4</td>\n",
       "      <td>1.5</td>\n",
       "      <td>0.4</td>\n",
       "      <td>Iris-setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>129</th>\n",
       "      <td>130</td>\n",
       "      <td>7.2</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.8</td>\n",
       "      <td>1.6</td>\n",
       "      <td>Iris-virginica</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>119</td>\n",
       "      <td>7.7</td>\n",
       "      <td>2.6</td>\n",
       "      <td>6.9</td>\n",
       "      <td>2.3</td>\n",
       "      <td>Iris-virginica</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>68</td>\n",
       "      <td>5.8</td>\n",
       "      <td>2.7</td>\n",
       "      <td>4.1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Iris-versicolor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.6</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>Iris-setosa</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Id  SepalLengthCm  SepalWidthCm  PetalLengthCm  PetalWidthCm  \\\n",
       "15    16            5.7           4.4            1.5           0.4   \n",
       "129  130            7.2           3.0            5.8           1.6   \n",
       "118  119            7.7           2.6            6.9           2.3   \n",
       "67    68            5.8           2.7            4.1           1.0   \n",
       "4      5            5.0           3.6            1.4           0.2   \n",
       "\n",
       "             Species  \n",
       "15       Iris-setosa  \n",
       "129   Iris-virginica  \n",
       "118   Iris-virginica  \n",
       "67   Iris-versicolor  \n",
       "4        Iris-setosa  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# load iris database\n",
    "data = pd.read_csv('C:\\\\Anaconda2\\\\books\\\\Iris.csv')\n",
    "data.sample(n=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Id</th>\n",
       "      <th>SepalLengthCm</th>\n",
       "      <th>SepalWidthCm</th>\n",
       "      <th>PetalLengthCm</th>\n",
       "      <th>PetalWidthCm</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>150.000000</td>\n",
       "      <td>150.000000</td>\n",
       "      <td>150.000000</td>\n",
       "      <td>150.000000</td>\n",
       "      <td>150.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>75.500000</td>\n",
       "      <td>5.843333</td>\n",
       "      <td>3.054000</td>\n",
       "      <td>3.758667</td>\n",
       "      <td>1.198667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>43.445368</td>\n",
       "      <td>0.828066</td>\n",
       "      <td>0.433594</td>\n",
       "      <td>1.764420</td>\n",
       "      <td>0.763161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>4.300000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.100000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>38.250000</td>\n",
       "      <td>5.100000</td>\n",
       "      <td>2.800000</td>\n",
       "      <td>1.600000</td>\n",
       "      <td>0.300000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>75.500000</td>\n",
       "      <td>5.800000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>4.350000</td>\n",
       "      <td>1.300000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>112.750000</td>\n",
       "      <td>6.400000</td>\n",
       "      <td>3.300000</td>\n",
       "      <td>5.100000</td>\n",
       "      <td>1.800000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>150.000000</td>\n",
       "      <td>7.900000</td>\n",
       "      <td>4.400000</td>\n",
       "      <td>6.900000</td>\n",
       "      <td>2.500000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Id  SepalLengthCm  SepalWidthCm  PetalLengthCm  PetalWidthCm\n",
       "count  150.000000     150.000000    150.000000     150.000000    150.000000\n",
       "mean    75.500000       5.843333      3.054000       3.758667      1.198667\n",
       "std     43.445368       0.828066      0.433594       1.764420      0.763161\n",
       "min      1.000000       4.300000      2.000000       1.000000      0.100000\n",
       "25%     38.250000       5.100000      2.800000       1.600000      0.300000\n",
       "50%     75.500000       5.800000      3.000000       4.350000      1.300000\n",
       "75%    112.750000       6.400000      3.300000       5.100000      1.800000\n",
       "max    150.000000       7.900000      4.400000       6.900000      2.500000"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SepalLengthCm</th>\n",
       "      <th>SepalWidthCm</th>\n",
       "      <th>PetalLengthCm</th>\n",
       "      <th>PetalWidthCm</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>86</th>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.458333</td>\n",
       "      <td>0.627119</td>\n",
       "      <td>0.583333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.222222</td>\n",
       "      <td>0.625000</td>\n",
       "      <td>0.067797</td>\n",
       "      <td>0.083333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>0.416667</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>0.508475</td>\n",
       "      <td>0.458333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>90</th>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>0.576271</td>\n",
       "      <td>0.458333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>0.916667</td>\n",
       "      <td>0.416667</td>\n",
       "      <td>0.949153</td>\n",
       "      <td>0.833333</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     SepalLengthCm  SepalWidthCm  PetalLengthCm  PetalWidthCm\n",
       "86        0.666667      0.458333       0.627119      0.583333\n",
       "17        0.222222      0.625000       0.067797      0.083333\n",
       "92        0.416667      0.250000       0.508475      0.458333\n",
       "90        0.333333      0.250000       0.576271      0.458333\n",
       "105       0.916667      0.416667       0.949153      0.833333"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_norm = data[['SepalLengthCm', 'SepalWidthCm', 'PetalLengthCm', 'PetalWidthCm']].apply(lambda x: (x - x.min()) / (x.max() - x.min()))\n",
    "df_norm.sample(n=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SepalLengthCm</th>\n",
       "      <th>SepalWidthCm</th>\n",
       "      <th>PetalLengthCm</th>\n",
       "      <th>PetalWidthCm</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>150.000000</td>\n",
       "      <td>150.000000</td>\n",
       "      <td>150.000000</td>\n",
       "      <td>150.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.428704</td>\n",
       "      <td>0.439167</td>\n",
       "      <td>0.467571</td>\n",
       "      <td>0.457778</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.230018</td>\n",
       "      <td>0.180664</td>\n",
       "      <td>0.299054</td>\n",
       "      <td>0.317984</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.222222</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.101695</td>\n",
       "      <td>0.083333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.416667</td>\n",
       "      <td>0.416667</td>\n",
       "      <td>0.567797</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.583333</td>\n",
       "      <td>0.541667</td>\n",
       "      <td>0.694915</td>\n",
       "      <td>0.708333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       SepalLengthCm  SepalWidthCm  PetalLengthCm  PetalWidthCm\n",
       "count     150.000000    150.000000     150.000000    150.000000\n",
       "mean        0.428704      0.439167       0.467571      0.457778\n",
       "std         0.230018      0.180664       0.299054      0.317984\n",
       "min         0.000000      0.000000       0.000000      0.000000\n",
       "25%         0.222222      0.333333       0.101695      0.083333\n",
       "50%         0.416667      0.416667       0.567797      0.500000\n",
       "75%         0.583333      0.541667       0.694915      0.708333\n",
       "max         1.000000      1.000000       1.000000      1.000000"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_norm.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Convert the Species labels to indexes for use with neural network.\n",
    "Iris-setoso = 0\n",
    "Iris-versicolor = 1\n",
    "Iris-virginica = 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Species</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>76</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>135</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Species\n",
       "76         1\n",
       "26         0\n",
       "125        2\n",
       "9          0\n",
       "135        2"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "target = data[['Species']].replace(['Iris-setosa','Iris-versicolor','Iris-virginica'],[0,1,2])\n",
    "target.sample(n=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SepalLengthCm</th>\n",
       "      <th>SepalWidthCm</th>\n",
       "      <th>PetalLengthCm</th>\n",
       "      <th>PetalWidthCm</th>\n",
       "      <th>Species</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.208333</td>\n",
       "      <td>0.508475</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>0.527778</td>\n",
       "      <td>0.375000</td>\n",
       "      <td>0.559322</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.627119</td>\n",
       "      <td>0.541667</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>0.361111</td>\n",
       "      <td>0.375000</td>\n",
       "      <td>0.440678</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>0.388889</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.593220</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    SepalLengthCm  SepalWidthCm  PetalLengthCm  PetalWidthCm  Species\n",
       "89       0.333333      0.208333       0.508475      0.500000        1\n",
       "97       0.527778      0.375000       0.559322      0.500000        1\n",
       "50       0.750000      0.500000       0.627119      0.541667        1\n",
       "64       0.361111      0.375000       0.440678      0.500000        1\n",
       "55       0.388889      0.333333       0.593220      0.500000        1"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.concat([df_norm, target], axis=1)\n",
    "df.sample(n=5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Mark some of the data for testing purpose.\n",
    "We'll test our network on unseen data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SepalLengthCm</th>\n",
       "      <th>SepalWidthCm</th>\n",
       "      <th>PetalLengthCm</th>\n",
       "      <th>PetalWidthCm</th>\n",
       "      <th>Species</th>\n",
       "      <th>train</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>144</th>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.541667</td>\n",
       "      <td>0.796610</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.208333</td>\n",
       "      <td>0.661017</td>\n",
       "      <td>0.583333</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128</th>\n",
       "      <td>0.583333</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.779661</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>2</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.541667</td>\n",
       "      <td>0.627119</td>\n",
       "      <td>0.625000</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.125000</td>\n",
       "      <td>0.576271</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     SepalLengthCm  SepalWidthCm  PetalLengthCm  PetalWidthCm  Species  train\n",
       "144       0.666667      0.541667       0.796610      1.000000        2   True\n",
       "72        0.555556      0.208333       0.661017      0.583333        1   True\n",
       "128       0.583333      0.333333       0.779661      0.833333        2  False\n",
       "56        0.555556      0.541667       0.627119      0.625000        1   True\n",
       "87        0.555556      0.125000       0.576271      0.500000        1   True"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_test_per = 90/100.0\n",
    "df['train'] = np.random.rand(len(df)) < train_test_per\n",
    "df.sample(n=5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Separate train and test Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SepalLengthCm</th>\n",
       "      <th>SepalWidthCm</th>\n",
       "      <th>PetalLengthCm</th>\n",
       "      <th>PetalWidthCm</th>\n",
       "      <th>Species</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>0.111111</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.101695</td>\n",
       "      <td>0.041667</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>0.527778</td>\n",
       "      <td>0.375000</td>\n",
       "      <td>0.559322</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>138</th>\n",
       "      <td>0.472222</td>\n",
       "      <td>0.416667</td>\n",
       "      <td>0.644068</td>\n",
       "      <td>0.708333</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>0.250000</td>\n",
       "      <td>0.625000</td>\n",
       "      <td>0.084746</td>\n",
       "      <td>0.041667</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>0.194444</td>\n",
       "      <td>0.416667</td>\n",
       "      <td>0.101695</td>\n",
       "      <td>0.041667</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     SepalLengthCm  SepalWidthCm  PetalLengthCm  PetalWidthCm  Species\n",
       "29        0.111111      0.500000       0.101695      0.041667        0\n",
       "97        0.527778      0.375000       0.559322      0.500000        1\n",
       "138       0.472222      0.416667       0.644068      0.708333        2\n",
       "27        0.250000      0.625000       0.084746      0.041667        0\n",
       "25        0.194444      0.416667       0.101695      0.041667        0"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train = df[df.train == 1]\n",
    "train = train.drop('train', axis=1).sample(frac=1)\n",
    "train.sample(n=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SepalLengthCm</th>\n",
       "      <th>SepalWidthCm</th>\n",
       "      <th>PetalLengthCm</th>\n",
       "      <th>PetalWidthCm</th>\n",
       "      <th>Species</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>0.222222</td>\n",
       "      <td>0.208333</td>\n",
       "      <td>0.338983</td>\n",
       "      <td>0.416667</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65</th>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.458333</td>\n",
       "      <td>0.576271</td>\n",
       "      <td>0.541667</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.208333</td>\n",
       "      <td>0.677966</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128</th>\n",
       "      <td>0.583333</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.779661</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>0.166667</td>\n",
       "      <td>0.458333</td>\n",
       "      <td>0.084746</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     SepalLengthCm  SepalWidthCm  PetalLengthCm  PetalWidthCm  Species\n",
       "98        0.222222      0.208333       0.338983      0.416667        1\n",
       "65        0.666667      0.458333       0.576271      0.541667        1\n",
       "146       0.555556      0.208333       0.677966      0.750000        2\n",
       "128       0.583333      0.333333       0.779661      0.833333        2\n",
       "34        0.166667      0.458333       0.084746      0.000000        0"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test = df[df.train == 0]\n",
    "test = test.drop('train', axis=1)\n",
    "test.sample(n=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.86111111,  0.33333333,  0.86440678,  0.75      ],\n",
       "       [ 0.61111111,  0.5       ,  0.69491525,  0.79166667],\n",
       "       [ 0.58333333,  0.5       ,  0.59322034,  0.58333333],\n",
       "       [ 0.22222222,  0.625     ,  0.06779661,  0.04166667],\n",
       "       [ 0.13888889,  0.45833333,  0.10169492,  0.04166667]])"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = train.values[:,:4]\n",
    "X[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 1, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0]])"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "targets = [[1,0,0],[0,1,0],[0,0,1]]\n",
    "y = np.array([targets[int(x)] for x in train.values[:,4:5]])\n",
    "y[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create backpropagating neural network\n",
    "Create 3 layers: Input, hidden and Output.\n",
    "Inputs = length and widths of the species\n",
    "Output = 3 values, each one indicating a species. ie Values 1, 0, 0 for the output indicates Iris-setosa\n",
    "w1 is a matrices of weight connecting Input and the hidden layer. Each node in input layer connects to each node in the hidden layer.\n",
    "\n",
    "Weight are randomized between -1 and 1."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4\n",
      "[[ 0.93405968  0.0944645   0.94536872  0.42963199  0.39545765]\n",
      " [-0.56782101  0.95254891 -0.98753949 -0.49403528 -0.13041694]\n",
      " [ 0.55876584 -0.60462985  0.72598647  0.96680135 -0.67231552]\n",
      " [ 0.19466789 -0.9820278  -0.22685743 -0.91167988  0.91330594]]\n"
     ]
    }
   ],
   "source": [
    "num_inputs = len(X[0])\n",
    "print(num_inputs)\n",
    "hidden_layer_neurons = 5\n",
    "np.random.seed(4)\n",
    "w1 = 2*np.random.random((num_inputs, hidden_layer_neurons)) - 1\n",
    "print(w1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "w2 are the weights of connections between hidden layer and output layer.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3\n",
      "[[-0.12770671  0.89795461  0.57261197]\n",
      " [ 0.7325786  -0.65366916 -0.85010283]\n",
      " [ 0.20148544 -0.66405563  0.46676034]\n",
      " [-0.18311228  0.05581765  0.87514317]\n",
      " [ 0.04339224 -0.78361323 -0.68355319]]\n"
     ]
    }
   ],
   "source": [
    "num_outputs = len(y[0])\n",
    "print(num_outputs)\n",
    "w2 = 2*np.random.random((hidden_layer_neurons, num_outputs)) - 1\n",
    "print(w2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Train the network by updating the weights using backpropogation.\n",
    "This is the crux of the network. The layers are fed forward using sigmoid activation function. The weighs are then updated based on error using gradient descent.\n",
    "Forward Propagation ( use current weights to caluculate output ):\n",
    "> node activation = output from previous layer (network inputs in case of first layer) * weights\n",
    "> node output = sigmoid activation function = 1 / ( 1 + exp( node activation ) )\n",
    "\n",
    "Backpropagation ( update network weights ):\n",
    "Error calculation ( how far off we are from the expected values ):\n",
    "> derivative (different for different activation functions) = output * ( 1 - output )\n",
    "> error (for the last layer) = ( expected - output ) * derivative\n",
    "> error (for other layers) = ( error calulated previously * that layer's weight ) * derivative\n",
    "Update weight based on error caculated:\n",
    "> Weight = weight + ( output * error * learning rate )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# taken from> https://gist.github.com/craffel/2d727968c3aaebd10359\n",
    "def draw_neural_net(ax, left, right, bottom, top, layer_sizes):\n",
    "    '''\n",
    "    Draw a neural network cartoon using matplotilb.\n",
    "    \n",
    "    :usage:\n",
    "        >>> fig = plt.figure(figsize=(12, 12))\n",
    "        >>> draw_neural_net(fig.gca(), .1, .9, .1, .9, [4, 7, 2])\n",
    "    \n",
    "    :parameters:\n",
    "        - ax : matplotlib.axes.AxesSubplot\n",
    "            The axes on which to plot the cartoon (get e.g. by plt.gca())\n",
    "        - left : float\n",
    "            The center of the leftmost node(s) will be placed here\n",
    "        - right : float\n",
    "            The center of the rightmost node(s) will be placed here\n",
    "        - bottom : float\n",
    "            The center of the bottommost node(s) will be placed here\n",
    "        - top : float\n",
    "            The center of the topmost node(s) will be placed here\n",
    "        - layer_sizes : list of int\n",
    "            List of layer sizes, including input and output dimensionality\n",
    "    '''\n",
    "    n_layers = len(layer_sizes)\n",
    "    v_spacing = (top - bottom)/float(max(layer_sizes))\n",
    "    h_spacing = (right - left)/float(len(layer_sizes) - 1)\n",
    "    # Nodes\n",
    "    for n, layer_size in enumerate(layer_sizes):\n",
    "        layer_top = v_spacing*(layer_size - 1)/2. + (top + bottom)/2.\n",
    "        for m in range(layer_size):\n",
    "            circle = plt.Circle((n*h_spacing + left, layer_top - m*v_spacing), v_spacing/4.,\n",
    "                                color='w', ec='k', zorder=4)\n",
    "            ax.add_artist(circle)\n",
    "    # Edges\n",
    "    for n, (layer_size_a, layer_size_b) in enumerate(zip(layer_sizes[:-1], layer_sizes[1:])):\n",
    "        layer_top_a = v_spacing*(layer_size_a - 1)/2. + (top + bottom)/2.\n",
    "        layer_top_b = v_spacing*(layer_size_b - 1)/2. + (top + bottom)/2.\n",
    "        for m in range(layer_size_a):\n",
    "            for o in range(layer_size_b):\n",
    "                line = plt.Line2D([n*h_spacing + left, (n + 1)*h_spacing + left],\n",
    "                                  [layer_top_a - m*v_spacing, layer_top_b - o*v_spacing], c='k')\n",
    "                ax.add_artist(line)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A Graphical representations of our network will be something like below\n",
    "The first set of 4 nodes is the input.\n",
    "The second set of 5 nodes is the hidden layer. \n",
    "The last set of 3 nodes is the output layer.\n",
    "\n",
    "All the nodes of a layer are fully connected to all nodes of the next layer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsgAAAK+CAYAAAC/0OiqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nlJdKgONyAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ffedd8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(12, 12))\n",
    "ax = fig.gca()\n",
    "ax.axis('off')\n",
    "draw_neural_net(ax, .1, .9, .1, .9, [4, 5, 3])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The sigmoid activation function squashes the input values between 0 and 1. This provides a consistant way for the network to deal with outputs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x97b0c18>]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEACAYAAACj0I2EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEZGE+P/BIVSrvzVk9AAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x9690ac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# sigmoid function representation\n",
    "_x = np.linspace( -5, 5, 50 )\n",
    "_y = 1 / ( 1 + np.exp( -_x ) )\n",
    "plt.plot( _x, _y )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('Error:', 0.044869912787412924)\n"
     ]
    }
   ],
   "source": [
    "learning_rate = 0.2 # slowly update the network\n",
    "for epoch in range(50000):\n",
    "    l1 = 1/(1 + np.exp(-(np.dot(X, w1)))) # sigmoid function\n",
    "    l2 = 1/(1 + np.exp(-(np.dot(l1, w2))))\n",
    "    er = (abs(y - l2)).mean()\n",
    "    l2_delta = (y - l2)*(l2 * (1-l2))\n",
    "    l1_delta = l2_delta.dot(w2.T) * (l1 * (1-l1))\n",
    "    w2 += l1.T.dot(l2_delta) * learning_rate\n",
    "    w1 += X.T.dot(l1_delta) * learning_rate\n",
    "print('Error:', er)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Test the network for accuracy.\n",
    "Run the network with the updated weights from training."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 1.   ,  0.   ,  0.   ],\n",
       "       [ 1.   ,  0.   ,  0.   ],\n",
       "       [ 1.   ,  0.   ,  0.   ],\n",
       "       [ 1.   ,  0.   ,  0.   ],\n",
       "       [ 1.   ,  0.   ,  0.   ],\n",
       "       [ 1.   ,  0.   ,  0.   ],\n",
       "       [ 1.   ,  0.   ,  0.   ],\n",
       "       [ 0.   ,  0.931,  0.044],\n",
       "       [ 0.   ,  1.   ,  0.   ],\n",
       "       [ 0.   ,  0.079,  0.937],\n",
       "       [ 0.   ,  1.   ,  0.   ],\n",
       "       [ 0.   ,  0.614,  0.344],\n",
       "       [ 0.   ,  0.998,  0.001],\n",
       "       [ 0.   ,  0.036,  0.975],\n",
       "       [ 0.   ,  0.039,  0.973],\n",
       "       [ 0.   ,  0.106,  0.913]])"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = test.values[:,:4]\n",
    "y = np.array([targets[int(x)] for x in test.values[:,4:5]])\n",
    "\n",
    "l1 = 1/(1 + np.exp(-(np.dot(X, w1))))\n",
    "l2 = 1/(1 + np.exp(-(np.dot(l1, w2))))\n",
    "\n",
    "np.round(l2,3)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the above maxtrix we take the maximum value (per row), which forms our predictions..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('The correct values is ', 0)\n",
      "             Species       Prediction\n",
      "2        Iris-setosa      Iris-setosa\n",
      "21       Iris-setosa      Iris-setosa\n",
      "22       Iris-setosa      Iris-setosa\n",
      "32       Iris-setosa      Iris-setosa\n",
      "34       Iris-setosa      Iris-setosa\n",
      "38       Iris-setosa      Iris-setosa\n",
      "48       Iris-setosa      Iris-setosa\n",
      "65   Iris-versicolor  Iris-versicolor\n",
      "66   Iris-versicolor  Iris-versicolor\n",
      "83   Iris-versicolor   Iris-virginica\n",
      "85   Iris-versicolor  Iris-versicolor\n",
      "86   Iris-versicolor  Iris-versicolor\n",
      "98   Iris-versicolor  Iris-versicolor\n",
      "115   Iris-virginica   Iris-virginica\n",
      "128   Iris-virginica   Iris-virginica\n",
      "146   Iris-virginica   Iris-virginica\n",
      "('Correct:', 15, '/', 16, ':', 0, '%')\n"
     ]
    }
   ],
   "source": [
    "yp = np.argmax(l2, axis=1) # prediction\n",
    "res = yp == np.argmax(y, axis=1)\n",
    "correct = np.sum(res)/len(res)\n",
    "print(\"The correct values is \",correct)\n",
    "\n",
    "testres = test[['Species']].replace([0,1,2], ['Iris-setosa','Iris-versicolor','Iris-virginica'])\n",
    "\n",
    "testres['Prediction'] = yp\n",
    "testres['Prediction'] = testres['Prediction'].replace([0,1,2], ['Iris-setosa','Iris-versicolor','Iris-virginica'])\n",
    "\n",
    "print(testres)\n",
    "print('Correct:',sum(res),'/',len(res), ':', (correct*100),'%')\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}