SoftmaxClassification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Softmax Classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "x_names = ['SepLength', 'SepWidth', 'PetLength', 'PetWidth']\n",
    "y_names = ['Prop_Iris-setosa', 'Prop_Iris-versicolor', 'Prop_Iris-virginica']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Reads inputs\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def inputs(file):\n",
    "    data = pd.read_table(filepath_or_buffer=file, delimiter=\",\", header=None, names=x_names + ['Prop'])\n",
    "    data = pd.get_dummies(data)\n",
    "    train = data.sample(frac=0.8)\n",
    "    test = data.drop(train.index)\n",
    "    return train, test\n",
    "\n",
    "\n",
    "Train, Test = inputs('data/iris.data.txt')"
   ]
  },
  {
   "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>SepLength</th>\n",
       "      <th>SepWidth</th>\n",
       "      <th>PetLength</th>\n",
       "      <th>PetWidth</th>\n",
       "      <th>Prop_Iris-setosa</th>\n",
       "      <th>Prop_Iris-versicolor</th>\n",
       "      <th>Prop_Iris-virginica</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>120.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>120.00000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>120.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>5.815000</td>\n",
       "      <td>3.060000</td>\n",
       "      <td>3.716667</td>\n",
       "      <td>1.17750</td>\n",
       "      <td>0.341667</td>\n",
       "      <td>0.325000</td>\n",
       "      <td>0.333333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.840933</td>\n",
       "      <td>0.417153</td>\n",
       "      <td>1.755775</td>\n",
       "      <td>0.76559</td>\n",
       "      <td>0.476257</td>\n",
       "      <td>0.470339</td>\n",
       "      <td>0.473381</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>4.300000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.10000</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>5.100000</td>\n",
       "      <td>2.800000</td>\n",
       "      <td>1.575000</td>\n",
       "      <td>0.20000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>5.700000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>4.250000</td>\n",
       "      <td>1.30000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>6.400000</td>\n",
       "      <td>3.325000</td>\n",
       "      <td>5.100000</td>\n",
       "      <td>1.80000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>7.900000</td>\n",
       "      <td>4.200000</td>\n",
       "      <td>6.900000</td>\n",
       "      <td>2.50000</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": [
       "        SepLength    SepWidth   PetLength   PetWidth  Prop_Iris-setosa  \\\n",
       "count  120.000000  120.000000  120.000000  120.00000        120.000000   \n",
       "mean     5.815000    3.060000    3.716667    1.17750          0.341667   \n",
       "std      0.840933    0.417153    1.755775    0.76559          0.476257   \n",
       "min      4.300000    2.000000    1.000000    0.10000          0.000000   \n",
       "25%      5.100000    2.800000    1.575000    0.20000          0.000000   \n",
       "50%      5.700000    3.000000    4.250000    1.30000          0.000000   \n",
       "75%      6.400000    3.325000    5.100000    1.80000          1.000000   \n",
       "max      7.900000    4.200000    6.900000    2.50000          1.000000   \n",
       "\n",
       "       Prop_Iris-versicolor  Prop_Iris-virginica  \n",
       "count            120.000000           120.000000  \n",
       "mean               0.325000             0.333333  \n",
       "std                0.470339             0.473381  \n",
       "min                0.000000             0.000000  \n",
       "25%                0.000000             0.000000  \n",
       "50%                0.000000             0.000000  \n",
       "75%                1.000000             1.000000  \n",
       "max                1.000000             1.000000  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Train.describe()"
   ]
  },
  {
   "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>SepLength</th>\n",
       "      <th>SepWidth</th>\n",
       "      <th>PetLength</th>\n",
       "      <th>PetWidth</th>\n",
       "      <th>Prop_Iris-setosa</th>\n",
       "      <th>Prop_Iris-versicolor</th>\n",
       "      <th>Prop_Iris-virginica</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>30.000000</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>30.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>5.956667</td>\n",
       "      <td>3.030000</td>\n",
       "      <td>3.926667</td>\n",
       "      <td>1.283333</td>\n",
       "      <td>0.300000</td>\n",
       "      <td>0.366667</td>\n",
       "      <td>0.333333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.777566</td>\n",
       "      <td>0.501137</td>\n",
       "      <td>1.819044</td>\n",
       "      <td>0.760256</td>\n",
       "      <td>0.466092</td>\n",
       "      <td>0.490133</td>\n",
       "      <td>0.479463</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>4.500000</td>\n",
       "      <td>2.200000</td>\n",
       "      <td>1.200000</td>\n",
       "      <td>0.200000</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>5.250000</td>\n",
       "      <td>2.750000</td>\n",
       "      <td>1.600000</td>\n",
       "      <td>0.400000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>6.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>4.400000</td>\n",
       "      <td>1.400000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>6.475000</td>\n",
       "      <td>3.275000</td>\n",
       "      <td>5.250000</td>\n",
       "      <td>1.800000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>7.600000</td>\n",
       "      <td>4.400000</td>\n",
       "      <td>6.600000</td>\n",
       "      <td>2.500000</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": [
       "       SepLength   SepWidth  PetLength   PetWidth  Prop_Iris-setosa  \\\n",
       "count  30.000000  30.000000  30.000000  30.000000         30.000000   \n",
       "mean    5.956667   3.030000   3.926667   1.283333          0.300000   \n",
       "std     0.777566   0.501137   1.819044   0.760256          0.466092   \n",
       "min     4.500000   2.200000   1.200000   0.200000          0.000000   \n",
       "25%     5.250000   2.750000   1.600000   0.400000          0.000000   \n",
       "50%     6.000000   3.000000   4.400000   1.400000          0.000000   \n",
       "75%     6.475000   3.275000   5.250000   1.800000          1.000000   \n",
       "max     7.600000   4.400000   6.600000   2.500000          1.000000   \n",
       "\n",
       "       Prop_Iris-versicolor  Prop_Iris-virginica  \n",
       "count             30.000000            30.000000  \n",
       "mean               0.366667             0.333333  \n",
       "std                0.490133             0.479463  \n",
       "min                0.000000             0.000000  \n",
       "25%                0.000000             0.000000  \n",
       "50%                0.000000             0.000000  \n",
       "75%                1.000000             1.000000  \n",
       "max                1.000000             1.000000  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Test.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Predicts Y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def inference(x, W, b):\n",
    "    return tf.nn.softmax(tf.matmul(x, W) + b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = Train[x_names]\n",
    "Y_train = Train[y_names]\n",
    "X_test = Test[x_names]\n",
    "Y_test = Test[y_names]\n",
    "x = tf.placeholder(tf.float32, [None, 4])\n",
    "W = tf.Variable(tf.zeros([4, 3]))\n",
    "b = tf.Variable(tf.zeros([3]))\n",
    "predict = inference(x, W, b)\n",
    "y = tf.placeholder(tf.float32, [None, 3])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Crosss Entropy and Accuaracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "crossEntropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits_v2(labels=y, logits=predict))\n",
    "predictionCorrection = tf.equal(tf.argmax(predict, axis=1), tf.argmax(y, axis=1))\n",
    "accuracy = tf.reduce_mean(tf.cast(predictionCorrection, tf.float32))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Learning Rate:  0.001\n",
      "Accuracy:  0.33333334\n",
      "Learning Rate:  0.01\n",
      "Accuracy:  0.33333334\n",
      "Learning Rate:  0.1\n",
      "Accuracy:  0.33333334\n",
      "Learning Rate:  1.0\n",
      "Accuracy:  0.4\n"
     ]
    }
   ],
   "source": [
    "for learningRate in [0.001, 0.01, 0.1, 1.0]:\n",
    "    optimizer = tf.train.GradientDescentOptimizer(learningRate).minimize(crossEntropy)\n",
    "    \n",
    "    with tf.Session() as session:\n",
    "        tf.global_variables_initializer().run()\n",
    "        for epoch in range(3):\n",
    "            session.run(optimizer, feed_dict={x: X_train, y: Y_train})\n",
    "        acc = session.run(accuracy, feed_dict={x: X_test, y: Y_test})\n",
    "        print(\"Learning Rate: \", learningRate)\n",
    "        print(\"Accuracy: \", acc)\n",
    "        session.close()"
   ]
  }
 ],
 "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": 1
}