sessions/linear-discriminant-analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "# Linear discriminant analysis\n",
    "\n",
    "LDA tries to identify attributes that account for the most variance between classes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "import numpy as np, pandas as pd, matplotlib.pyplot as plt\n",
    "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis\n",
    "from utils import plot_decision"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "## Load and prep the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": 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>sepal_length</th>\n",
       "      <th>sepal_width</th>\n",
       "      <th>petal_length</th>\n",
       "      <th>petal_width</th>\n",
       "      <th>species</th>\n",
       "      <th>species_label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5.1</td>\n",
       "      <td>3.5</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>Iris-setosa</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4.9</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.4</td>\n",
       "      <td>0.2</td>\n",
       "      <td>Iris-setosa</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4.7</td>\n",
       "      <td>3.2</td>\n",
       "      <td>1.3</td>\n",
       "      <td>0.2</td>\n",
       "      <td>Iris-setosa</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.6</td>\n",
       "      <td>3.1</td>\n",
       "      <td>1.5</td>\n",
       "      <td>0.2</td>\n",
       "      <td>Iris-setosa</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\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",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   sepal_length  sepal_width  petal_length  petal_width      species  \\\n",
       "0           5.1          3.5           1.4          0.2  Iris-setosa   \n",
       "1           4.9          3.0           1.4          0.2  Iris-setosa   \n",
       "2           4.7          3.2           1.3          0.2  Iris-setosa   \n",
       "3           4.6          3.1           1.5          0.2  Iris-setosa   \n",
       "4           5.0          3.6           1.4          0.2  Iris-setosa   \n",
       "\n",
       "   species_label  \n",
       "0              0  \n",
       "1              0  \n",
       "2              0  \n",
       "3              0  \n",
       "4              0  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# load the iris data\n",
    "df = pd.read_csv('data/iris.csv')\n",
    "df['species_label'], uniques = pd.factorize(df['species'])\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "# select features\n",
    "y = df['species_label']\n",
    "X = df[['petal_length', 'petal_width', 'sepal_length', 'sepal_width']]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Reduce from 4 dimensions to 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "lda = LinearDiscriminantAnalysis(n_components=2)\n",
    "X_reduced = lda.fit(X, y).transform(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXkAAAEICAYAAAC6fYRZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1649ce1eef0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "for species_label, species_name in enumerate(uniques):\n",
    "    matches = np.array(y==species_label)\n",
    "    ax.scatter(x=X_reduced[matches, 0], \n",
    "               y=X_reduced[matches, 1], \n",
    "               alpha=0.8, \n",
    "               label=species_name)\n",
    "ax.legend(loc='best', scatterpoints=1)\n",
    "ax.set_title('LDA of IRIS dataset')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "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": "3.5.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}