Mehraveh lin reg.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.preprocessing import LabelEncoder, OneHotEncoder\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "offc = pd.read_csv('data/toy.officer_data.csv',sep=',')\n",
    "cmpl = pd.read_csv('data/toy.complaint_data.csv', sep=',')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>first.name</th>\n",
       "      <th>last.name</th>\n",
       "      <th>appointed.date</th>\n",
       "      <th>race</th>\n",
       "      <th>gender</th>\n",
       "      <th>birth.year</th>\n",
       "      <th>age</th>\n",
       "      <th>rank</th>\n",
       "      <th>primary</th>\n",
       "      <th>secondary</th>\n",
       "      <th>tertiary</th>\n",
       "      <th>officer_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>   KARINA</td>\n",
       "      <td>       AARON</td>\n",
       "      <td> 2005-09-26</td>\n",
       "      <td> HISPANIC</td>\n",
       "      <td> FEMALE</td>\n",
       "      <td> 1980</td>\n",
       "      <td> 35</td>\n",
       "      <td> POLICE OFFICER</td>\n",
       "      <td> 0</td>\n",
       "      <td> 1</td>\n",
       "      <td> 0</td>\n",
       "      <td> 1139</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>     AZIZ</td>\n",
       "      <td> ABDELMAJEID</td>\n",
       "      <td> 2008-04-28</td>\n",
       "      <td>  A_OTHER</td>\n",
       "      <td>   MALE</td>\n",
       "      <td> 1984</td>\n",
       "      <td> 31</td>\n",
       "      <td> POLICE OFFICER</td>\n",
       "      <td> 0</td>\n",
       "      <td> 0</td>\n",
       "      <td> 1</td>\n",
       "      <td>  434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td> JENNIFER</td>\n",
       "      <td>     ACCARDO</td>\n",
       "      <td> 2004-11-29</td>\n",
       "      <td>    WHITE</td>\n",
       "      <td> FEMALE</td>\n",
       "      <td> 1979</td>\n",
       "      <td> 37</td>\n",
       "      <td> POLICE OFFICER</td>\n",
       "      <td> 0</td>\n",
       "      <td> 0</td>\n",
       "      <td> 1</td>\n",
       "      <td> 2515</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td> ROSEMARY</td>\n",
       "      <td>     ACCARDO</td>\n",
       "      <td> 1986-08-11</td>\n",
       "      <td>    WHITE</td>\n",
       "      <td> FEMALE</td>\n",
       "      <td> 1952</td>\n",
       "      <td> 63</td>\n",
       "      <td> POLICE OFFICER</td>\n",
       "      <td> 0</td>\n",
       "      <td> 0</td>\n",
       "      <td> 1</td>\n",
       "      <td> 1149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>   THOMAS</td>\n",
       "      <td>     ACCARDO</td>\n",
       "      <td> 2001-04-30</td>\n",
       "      <td>    WHITE</td>\n",
       "      <td>   MALE</td>\n",
       "      <td> 1967</td>\n",
       "      <td> 49</td>\n",
       "      <td> POLICE OFFICER</td>\n",
       "      <td> 0</td>\n",
       "      <td> 1</td>\n",
       "      <td> 0</td>\n",
       "      <td> 2134</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  first.name    last.name appointed.date      race  gender  birth.year  age  \\\n",
       "0     KARINA        AARON     2005-09-26  HISPANIC  FEMALE        1980   35   \n",
       "1       AZIZ  ABDELMAJEID     2008-04-28   A_OTHER    MALE        1984   31   \n",
       "2   JENNIFER      ACCARDO     2004-11-29     WHITE  FEMALE        1979   37   \n",
       "3   ROSEMARY      ACCARDO     1986-08-11     WHITE  FEMALE        1952   63   \n",
       "4     THOMAS      ACCARDO     2001-04-30     WHITE    MALE        1967   49   \n",
       "\n",
       "             rank  primary  secondary  tertiary  officer_id  \n",
       "0  POLICE OFFICER        0          1         0        1139  \n",
       "1  POLICE OFFICER        0          0         1         434  \n",
       "2  POLICE OFFICER        0          0         1        2515  \n",
       "3  POLICE OFFICER        0          0         1        1149  \n",
       "4  POLICE OFFICER        0          1         0        2134  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "offc.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>crid</th>\n",
       "      <th>incident_date</th>\n",
       "      <th>beat_2012_geocoded</th>\n",
       "      <th>complaintcategory</th>\n",
       "      <th>finalfinding</th>\n",
       "      <th>officer_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td> 1022797</td>\n",
       "      <td> 01-Jan-2009  0:08</td>\n",
       "      <td>NaN</td>\n",
       "      <td> 03G-MISCELLANEOUS</td>\n",
       "      <td> EX</td>\n",
       "      <td> 1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td> 1022797</td>\n",
       "      <td> 01-Jan-2009  0:08</td>\n",
       "      <td>NaN</td>\n",
       "      <td> 03G-MISCELLANEOUS</td>\n",
       "      <td> EX</td>\n",
       "      <td> 2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td> 1022797</td>\n",
       "      <td> 01-Jan-2009  0:08</td>\n",
       "      <td>NaN</td>\n",
       "      <td> 03G-MISCELLANEOUS</td>\n",
       "      <td> EX</td>\n",
       "      <td> 3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td> 1022797</td>\n",
       "      <td> 01-Jan-2009  0:08</td>\n",
       "      <td>NaN</td>\n",
       "      <td> 03G-MISCELLANEOUS</td>\n",
       "      <td> EX</td>\n",
       "      <td> 4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td> 1022797</td>\n",
       "      <td> 01-Jan-2009  0:08</td>\n",
       "      <td>NaN</td>\n",
       "      <td> 03G-MISCELLANEOUS</td>\n",
       "      <td> EX</td>\n",
       "      <td> 5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      crid      incident_date  beat_2012_geocoded  complaintcategory  \\\n",
       "0  1022797  01-Jan-2009  0:08                 NaN  03G-MISCELLANEOUS   \n",
       "1  1022797  01-Jan-2009  0:08                 NaN  03G-MISCELLANEOUS   \n",
       "2  1022797  01-Jan-2009  0:08                 NaN  03G-MISCELLANEOUS   \n",
       "3  1022797  01-Jan-2009  0:08                 NaN  03G-MISCELLANEOUS   \n",
       "4  1022797  01-Jan-2009  0:08                 NaN  03G-MISCELLANEOUS   \n",
       "\n",
       "  finalfinding  officer_id  \n",
       "0           EX           1  \n",
       "1           EX           2  \n",
       "2           EX           3  \n",
       "3           EX           4  \n",
       "4           EX           5  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cmpl.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X = offc\n",
    "X = X.drop(X.columns[[0,1,5,7,11]], axis=1)\n",
    "X[X.columns[0]] = X[X.columns[0]].astype(str).str[0:4].astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X_cat = X[X.columns[[1,2]]]\n",
    "Y=X_cat.apply(LabelEncoder().fit_transform)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X[X.columns[[1,2]]] = Y\n",
    "np.where(np.isnan(X))\n",
    "X.iat[np.where(np.isnan(X))[0][0],np.where(np.isnan(X))[1][0]]=0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_race = pd.get_dummies(X['race'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "add one-hot encoded races"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X = X.drop(X.columns[[1]], axis=1)\n",
    "X = X.join(X_race, how='outer')\n",
    "Y = cmpl.groupby('officer_id').count()['crid']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn import linear_model\n",
    "clf = linear_model.LinearRegression()\n",
    "X2=X[X.columns[[6,7,8,9]]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "clf.fit(X2,Y)\n",
    "predicts = clf.predict(X2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x10c943b50>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgYAAAFkCAYAAABb6/NsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10c840490>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "A = np.where(X2)[1]\n",
    "plt.scatter(Y,predicts, c=A,s=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from chris_rasmus_graph_features import *\n",
    "from count_lagged_complaints import *\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import LabelEncoder\n",
    "\n",
    "offc = pd.read_csv('/Users/Mehraveh/Desktop/policy_lab/toy.officer_data.csv',sep=',')\n",
    "cmpl = pd.read_csv('/Users/Mehraveh/Desktop/policy_lab/toy.complaint_data.csv', sep=',')\n",
    "\n",
    "r = add_complaints_by_year(2012, 4, cmpl, offc)\n",
    "\n",
    "Y = add_complaints_by_year(2015, 3, cmpl, offc)\n",
    "\n",
    "Y=Y[[Y.columns[Y.shape[1]-1],Y.columns[Y.shape[1]-2],Y.columns[Y.shape[1]-3]]]\n",
    "Y=Y.sum(axis=1)\n",
    "\n",
    "X = r\n",
    "X = X.drop(X.columns[[0,1,5,7,11]], axis=1)\n",
    "X[X.columns[0]] = X[X.columns[0]].astype(str).str[0:4].astype(int)\n",
    "X_cat = X[X.columns[[1,2]]]\n",
    "XX=X_cat.apply(LabelEncoder().fit_transform)\n",
    "\n",
    "X[X.columns[[1,2]]] = XX\n",
    "np.where(np.isnan(X))\n",
    "\n",
    "X_race = pd.get_dummies(X['race'])\n",
    "X_race.columns = ['race0','race1','race2','race3']\n",
    "\n",
    "X = X.drop(X.columns[[1]], axis=1)\n",
    "X = X.join(X_race, how='outer')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.623852917479\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXYAAAEACAYAAACnJV25AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1183d2b70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn import linear_model\n",
    "from sklearn.cross_validation import LeaveOneOut, cross_val_predict\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "clf = linear_model.LinearRegression()\n",
    "\n",
    "predicts = cross_val_predict(clf, X, Y, cv=3)\n",
    "\n",
    "plt.scatter(Y,predicts, s=100)\n",
    "mean_squared_error(Y, predicts)\n",
    "correlation = np.corrcoef(Y, predicts)[0,1]\n",
    "print(correlation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.603088645624\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXYAAAEACAYAAACnJV25AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11b41afd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.ensemble import GradientBoostingClassifier, GradientBoostingRegressor\n",
    "\n",
    "clf = GradientBoostingRegressor(n_estimators=500,learning_rate=0.05,min_samples_split=5,\n",
    "        min_samples_leaf=15,max_depth=5,max_features='sqrt',random_state=12,\n",
    "        subsample =0.8,verbose=0,warm_start = True)\n",
    "\n",
    "predicts = cross_val_predict(clf, X, Y, cv=3)\n",
    "\n",
    "#clf.fit(X2,Y)\n",
    "#predicts = clf.predict(X2)\n",
    "\n",
    "mean_squared_error(Y, predicts)\n",
    "plt.plot(Y,predicts,'*')\n",
    "correlation = np.corrcoef(Y, predicts)[0,1]\n",
    "print(correlation)\n"
   ]
  },
  {
   "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": "3.5.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}