DecisionTree.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Decison Trees"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we'll load some fake data on past hires I made up. Note how we use pandas to convert a csv file into a DataFrame:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn import tree\n",
    "\n",
    "input_file = \"e:/sundog-consult/udemy/datascience/PastHires.csv\"\n",
    "df = pd.read_csv(input_file, header = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>Years Experience</th>\n",
       "      <th>Employed?</th>\n",
       "      <th>Previous employers</th>\n",
       "      <th>Level of Education</th>\n",
       "      <th>Top-tier school</th>\n",
       "      <th>Interned</th>\n",
       "      <th>Hired</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>Y</td>\n",
       "      <td>4</td>\n",
       "      <td>BS</td>\n",
       "      <td>N</td>\n",
       "      <td>N</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>0</td>\n",
       "      <td>BS</td>\n",
       "      <td>Y</td>\n",
       "      <td>Y</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>7</td>\n",
       "      <td>N</td>\n",
       "      <td>6</td>\n",
       "      <td>BS</td>\n",
       "      <td>N</td>\n",
       "      <td>N</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>Y</td>\n",
       "      <td>1</td>\n",
       "      <td>MS</td>\n",
       "      <td>Y</td>\n",
       "      <td>N</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>20</td>\n",
       "      <td>N</td>\n",
       "      <td>2</td>\n",
       "      <td>PhD</td>\n",
       "      <td>Y</td>\n",
       "      <td>N</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Years Experience Employed?  Previous employers Level of Education  \\\n",
       "0                10         Y                   4                 BS   \n",
       "1                 0         N                   0                 BS   \n",
       "2                 7         N                   6                 BS   \n",
       "3                 2         Y                   1                 MS   \n",
       "4                20         N                   2                PhD   \n",
       "\n",
       "  Top-tier school Interned Hired  \n",
       "0               N        N     Y  \n",
       "1               Y        Y     Y  \n",
       "2               N        N     N  \n",
       "3               Y        N     Y  \n",
       "4               Y        N     N  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "scikit-learn needs everything to be numerical for decision trees to work. So, we'll map Y,N to 1,0 and levels of education to some scale of 0-2. In the real world, you'd need to think about how to deal with unexpected or missing data! By using map(), we know we'll get NaN for unexpected values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>Years Experience</th>\n",
       "      <th>Employed?</th>\n",
       "      <th>Previous employers</th>\n",
       "      <th>Level of Education</th>\n",
       "      <th>Top-tier school</th>\n",
       "      <th>Interned</th>\n",
       "      <th>Hired</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>20</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Years Experience  Employed?  Previous employers  Level of Education  \\\n",
       "0                10          1                   4                   0   \n",
       "1                 0          0                   0                   0   \n",
       "2                 7          0                   6                   0   \n",
       "3                 2          1                   1                   1   \n",
       "4                20          0                   2                   2   \n",
       "\n",
       "   Top-tier school  Interned  Hired  \n",
       "0                0         0      1  \n",
       "1                1         1      1  \n",
       "2                0         0      0  \n",
       "3                1         0      1  \n",
       "4                1         0      0  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d = {'Y': 1, 'N': 0}\n",
    "df['Hired'] = df['Hired'].map(d)\n",
    "df['Employed?'] = df['Employed?'].map(d)\n",
    "df['Top-tier school'] = df['Top-tier school'].map(d)\n",
    "df['Interned'] = df['Interned'].map(d)\n",
    "d = {'BS': 0, 'MS': 1, 'PhD': 2}\n",
    "df['Level of Education'] = df['Level of Education'].map(d)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next we need to separate the features from the target column that we're trying to bulid a decision tree for."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Years Experience',\n",
       " 'Employed?',\n",
       " 'Previous employers',\n",
       " 'Level of Education',\n",
       " 'Top-tier school',\n",
       " 'Interned']"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features = list(df.columns[:6])\n",
    "features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now actually construct the decision tree:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "y = df[\"Hired\"]\n",
    "X = df[features]\n",
    "clf = tree.DecisionTreeClassifier()\n",
    "clf = clf.fit(X,y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "... and display it. Note you need to have pyplot2 installed for this to work.\n",
    "\n",
    "To read this decision tree, each condition branches left for \"true\" and right for \"false\". When you end up at a value, the value array represents how many samples exist in each target value. So value = [0. 5.] mean there are 0 \"no hires\" and 5 \"hires\" by the tim we get to that point. value = [3. 0.] means 3 no-hires and 0 hires."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiUAAAJ8CAYAAAAh7NHUAAAABmJLR0QA/wD/AP+gvaeTAAAgAElE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\nrkJggg==\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import Image  \n",
    "from sklearn.externals.six import StringIO  \n",
    "import pydot \n",
    "\n",
    "dot_data = StringIO()  \n",
    "tree.export_graphviz(clf, out_file=dot_data,  \n",
    "                         feature_names=features)  \n",
    "graph = pydot.graph_from_dot_data(dot_data.getvalue())  \n",
    "Image(graph.create_png())  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ensemble learning: using a random forest"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We'll use a random forest of 10 decision trees to predict employment of specific candidate profiles:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1]\n",
      "[0]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "clf = RandomForestClassifier(n_estimators=10)\n",
    "clf = clf.fit(X, y)\n",
    "\n",
    "#Predict employment of an employed 10-year veteran\n",
    "print clf.predict([[10, 1, 4, 0, 0, 0]])\n",
    "#...and an unemployed 10-year veteran\n",
    "print clf.predict([[10, 0, 4, 0, 0, 0]])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Activity"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Modify the test data to create an alternate universe where everyone I hire everyone I normally wouldn't have, and vice versa. Compare the resulting decision tree to the one from the original data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": []
  }
 ],
 "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": "2.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}