Notebook 3 - KNN (Baseline).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## KNN "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import packages."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "16607"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "path_to_data = '/home/bsknight/Documents/Personal_Training_Git/Udacity/' + \\\n",
    "               'Udacity-Machine_Learning_Nanodegree/Capstone_Project/Data' \n",
    "os.chdir(path_to_data)\n",
    "df = pd.read_csv('./munged_df.csv', index_col='account_id')\n",
    "len(df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Extract features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "feature_cols = list(df.columns[:-1])\n",
    "target_col = df.columns[-1] \n",
    "X_all = df[feature_cols]\n",
    "y_all = df[target_col] "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Scale the features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import RobustScaler\n",
    "scaler = RobustScaler()\n",
    "X_all = scaler.fit_transform(X_all)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Split the data into training and testing sets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training set has 14946 samples.\n",
      "Testing set has 1661 samples.\n"
     ]
    }
   ],
   "source": [
    "from sklearn.cross_validation import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_all , \n",
    "                                                    y_all, \n",
    "                                                    stratify=y_all,\n",
    "                                                    test_size=0.1, \n",
    "                                                    random_state=1)\n",
    "print \"Training set has {} samples.\".format(X_train.shape[0])\n",
    "print \"Testing set has {} samples.\".format(X_test.shape[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create the KNN classifier."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "neigh = KNeighborsClassifier(n_jobs=-1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Deploy the KNN classifier."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "model = neigh.fit(X_train, y_train) \n",
    "y_pred = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Derive the initial F2 score."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.1635514018691589"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import fbeta_score\n",
    "fbeta_score(y_test, y_pred, beta=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create the custom scorer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics import fbeta_score, make_scorer\n",
    "ftwo_scorer = make_scorer(fbeta_score, beta=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/bsknight/anaconda3/envs/py2/lib/python2.7/site-packages/sklearn/metrics/classification.py:1074: UndefinedMetricWarning: F-score is ill-defined and being set to 0.0 due to no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0    0.04\n",
       "dtype: float64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.cross_validation import cross_val_score\n",
    "cv_results = pd.DataFrame(cross_val_score(model, X_test, y_test, cv=100, scoring=ftwo_scorer))\n",
    "cv_results.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    0.04\n",
       "dtype: float64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.cross_validation import cross_val_score\n",
    "cv_results = pd.DataFrame(cross_val_score(model, X_test, y_test, cv=100, scoring='recall'))\n",
    "cv_results.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/bsknight/anaconda3/envs/py2/lib/python2.7/site-packages/sklearn/metrics/classification.py:1074: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0    0.04\n",
       "dtype: float64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.cross_validation import cross_val_score\n",
    "cv_results = pd.DataFrame(cross_val_score(model, X_test, y_test, cv=100, scoring='precision'))\n",
    "cv_results.mean()"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda env:py2]",
   "language": "python",
   "name": "conda-env-py2-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}