Melbourne_Grants.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "#import xgboost\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "matplotlib.style.use('ggplot')\n",
    "matplotlib.interactive(True)\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import confusion_matrix, roc_curve, auc\n",
    "from sklearn import preprocessing\n",
    "\n",
    "\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.gaussian_process import GaussianProcessClassifier\n",
    "from sklearn.gaussian_process.kernels import RBF\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier, AdaBoostClassifier, GradientBoostingClassifier\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.discriminant_analysis import QuadraticDiscriminantAnalysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ramya/anaconda3/lib/python3.5/site-packages/IPython/core/interactiveshell.py:2717: DtypeWarning: Columns (90,105,119,120,123,124,132,134,135,138,139,147,149,150,153,154,162,164,165,168,169,177,179,183,184,192,194,198,199,207,209,213,214,224,237,239,244) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  interactivity=interactivity, compiler=compiler, result=result)\n"
     ]
    }
   ],
   "source": [
    "df = pd.read_csv(\"./data/unimelb_training.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df[\"Sponsor.Code\"] = df[\"Sponsor.Code\"].fillna(df[\"Sponsor.Code\"].value_counts().index[0])\n",
    "df = df.join(pd.get_dummies(df[\"Sponsor.Code\"], prefix = \"S\"))\n",
    "#df.drop([\"Sponsor.Code\"], axis = 1, inplace = True)\n",
    "\n",
    "\n",
    "df[\"Grant.Category.Code\"] = df[\"Grant.Category.Code\"].fillna(df[\n",
    "        \"Grant.Category.Code\"].value_counts().index[0])\n",
    "df = df.join(pd.get_dummies(df[\"Grant.Category.Code\"], prefix = \"G\"))\n",
    "#df.drop([\"Grant.Category.Code\"], axis = 1, inplace = True)\n",
    "\n",
    "\n",
    "\n",
    "df[\"Contract.Value.Band...see.note.A\"] = df[\"Contract.Value.Band...see.note.A\"].fillna(\n",
    "    df[\"Contract.Value.Band...see.note.A\"].value_counts().index[0])\n",
    "\n",
    "#df[\"Sponsor.Code\"] = df[\"Sponsor.Code\"].astype('category', ordered = False)\n",
    "\n",
    "#df[\"Grant.Category.Code\"] = df[\"Grant.Category.Code\"].astype('category', ordered = False)\n",
    "\n",
    "df[\"Start.date\"] = pd.to_datetime(df[\"Start.date\"])\n",
    "df[\"Start.date\"] = df[\"Start.date\"].astype('int64')//1e15"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df.drop([\"Sponsor.Code\"], axis = 1, inplace = True)\n",
    "df.drop([\"Grant.Category.Code\"], axis = 1, inplace = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "contract_value_ordered = ['A ', 'B ', 'C ', 'D ', 'E ', 'F ', 'G ', 'H ', 'I ', 'J ', 'K ', 'L ', 'M ',\n",
    "                         'N ', 'O ', 'P ', 'Q ']\n",
    "df['Contract.Value.Band...see.note.A'] = df[\"Contract.Value.Band...see.note.A\"].astype(\n",
    "    'category', ordered = True, categories = contract_value_ordered).cat.codes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "for feature in [\"RFCD.Code.\", \"RFCD.Percentage.\", \"SEO.Code.\", \"SEO.Percentage.\"]:\n",
    "    for i in range(1, 6):\n",
    "        df[feature+str(i)] = df[feature+str(i)].fillna(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "RFCD_dummy = pd.DataFrame()\n",
    "for i in range(1, 6):\n",
    "    RFCD_dummy = RFCD_dummy.add(pd.get_dummies(df[\"RFCD.Code.\" + str(i)],\n",
    "                                       columns = [\"RFCD.Code.\" + str(i)], prefix = \"RFCD\").multiply(\n",
    "                                       df[\"RFCD.Percentage.\" + str(i)],axis = 0), \n",
    "                        fill_value = 0.0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "SEO_dummy = pd.DataFrame()\n",
    "for i in range(1, 6):\n",
    "    SEO_dummy = SEO_dummy.add(pd.get_dummies(df[\"SEO.Code.\" + str(i)],\n",
    "                                       columns = [\"SEO.Code.\" + str(i)], prefix = \"SEO\").multiply(\n",
    "                                       df[\"SEO.Percentage.\" + str(i)],axis = 0), \n",
    "                        fill_value = 0.0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df2 = df.join(RFCD_dummy.join(SEO_dummy))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "for i in range(1,6):\n",
    "    df2.drop([\"RFCD.Code.\" + str(i), \"RFCD.Percentage.\" + str(i),\n",
    "                 \"SEO.Code.\" + str(i), \"SEO.Percentage.\" + str(i)],\n",
    "                   axis = 1, inplace = True)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "___\n",
    "** Droping all columns with NaN values (have to modify after data cleaning)**\n",
    "___"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df2.dropna(axis = 1, how = 'any', inplace = True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "___\n",
    "** After data cleansing for the remaining set of columns, please check this **\n",
    "___"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "y = df2[\"Grant.Status\"]\n",
    "X = df2.drop([\"Grant.Status\"], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X[\"Start.date\"] = X[\"Start.date\"].astype('int64')//1e15"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.3, random_state = 7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Forest Classifier Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "model = RandomForestClassifier(n_estimators = 45, max_depth = 15, \n",
    "                               min_samples_split = 4, min_samples_leaf = 2,\n",
    "                               oob_score = True, random_state = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=15, max_features='auto', max_leaf_nodes=None,\n",
       "            min_impurity_split=1e-07, min_samples_leaf=2,\n",
       "            min_samples_split=4, min_weight_fraction_leaf=0.0,\n",
       "            n_estimators=45, n_jobs=1, oob_score=True, random_state=1,\n",
       "            verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "score = model.oob_score_"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7133716160787531"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "score = model.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.72789896670493681"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Making predictions\n",
    "\n",
    "__Linear Regression__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ramya/anaconda3/lib/python3.5/site-packages/pandas/core/indexing.py:141: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  self._setitem_with_indexer(indexer, value)\n",
      "/media/ramya/D/Data_Science_Retreat/Jose_Minicompetition/DSR_minicompetition/clean_team_data.py:69: PerformanceWarning: dropping on a non-lexsorted multi-index without a level parameter may impact performance.\n",
      "  df.drop(\"Person.ID\", axis=1, inplace=True)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of columns in the first part is  1609\n",
      "Number of columns in the second part is  154\n",
      "Number of columns in the complete & cleaned data sets  is  1763\n"
     ]
    }
   ],
   "source": [
    "from combine_two_parts import clean_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "y = clean_data[\"Grant.Status\"]\n",
    "X = clean_data.drop([\"Grant.Status\"], axis = 1)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.3, random_state = 7)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check with different features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "names = [\n",
    "        \"Neural Net\", \n",
    "        #\"Logistic Regression\", \n",
    "        #\"Nearest Neighbors\", \n",
    "        #\"Linear SVM\", \n",
    "        #\"RBF SVM\", \n",
    "        #\"Gaussian Process\",\n",
    "        #\"Decision Tree\", \n",
    "        \"Random Forest\", \n",
    "        #\"AdaBoost\",\n",
    "        #\"Naive Bayes\", \n",
    "        #\"QDA\"\n",
    "        ]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "classifiers = [\n",
    "        MLPClassifier(alpha=0.001),\n",
    "        #LogisticRegression(C=1e-10, penalty='l2', tol=1e-6),\n",
    "        #KNeighborsClassifier(3),\n",
    "        #SVC(kernel=\"linear\", C=0.025),\n",
    "        #SVC(gamma=2, C=1),\n",
    "        #GaussianProcessClassifier(1.0 * RBF(1.0), warm_start=True),\n",
    "        #DecisionTreeClassifier(max_depth=5),\n",
    "        RandomForestClassifier(n_estimators = 45, max_depth = 15, \n",
    "                               min_samples_split = 4, min_samples_leaf = 2,\n",
    "                               oob_score = True, random_state = 1),\n",
    "        #AdaBoostClassifier(),\n",
    "        #GaussianNB(),\n",
    "        #QuadraticDiscriminantAnalysis()\n",
    "        ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Neural Net\n",
      "True positives: 0.5809367296631059\n",
      "False negatives: 0.419063270336894\n",
      "True negatives: 0.7005730659025788\n",
      "True positives: 0.2994269340974212\n",
      "\n",
      "ROC AUC: 0.64\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAioAAAGSCAYAAADNQeo2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f6b306ef898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random Forest\n",
      "True positives: 0.5891536565324569\n",
      "False negatives: 0.4108463434675431\n",
      "True negatives: 0.8638968481375359\n",
      "True positives: 0.1361031518624642\n",
      "\n",
      "ROC AUC: 0.73\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAioAAAGSCAYAAADNQeo2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nK5EnAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f6b306c4d68>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cms = {}\n",
    "n_neg = float(y_test.value_counts()[0])\n",
    "n_pos = float(y_test.value_counts()[1])\n",
    "for name, clf in zip(names, classifiers):\n",
    "    print(name)\n",
    "    clf.fit(X_train, y_train)\n",
    "    cm = confusion_matrix(y_test, clf.predict(X_test))\n",
    "    cms[name] = cm\n",
    "    print(\"True positives: {}\".format(cm[1,1] / n_pos))\n",
    "    print(\"False negatives: {}\".format(cm[1,0] / n_pos))\n",
    "    print(\"True negatives: {}\".format(cm[0,0] / n_neg))\n",
    "    print(\"True positives: {}\".format(cm[0,1] / n_neg))\n",
    "    print(\"\")\n",
    "    fpr, tpr, _ = roc_curve(y_test, clf.predict(X_test))\n",
    "    \n",
    "    roc_auc = auc(fpr, tpr)\n",
    "    print(\"ROC AUC: %0.2f\" % roc_auc)\n",
    "    \n",
    "    plt.figure()\n",
    "    plt.plot(fpr, tpr, label='ROC curve (area = %0.2f)' % roc_auc)\n",
    "    plt.plot([0, 1], [0, 1], 'k--')\n",
    "    plt.xlabel('False Positive Rate')\n",
    "    plt.ylabel('True Positive Rate')\n",
    "    plt.title('ROC Curve')\n",
    "    plt.legend(loc=\"lower right\")\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Neural Net\n"
     ]
    },
    {
     "ename": "AttributeError",
     "evalue": "'MLPClassifier' object has no attribute 'apply'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-27-40e6ee3d81ee>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      2\u001b[0m     \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mname\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m     \u001b[0mclf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX_train\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_train\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 4\u001b[0;31m     \u001b[0mclf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mapply\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m: 'MLPClassifier' object has no attribute 'apply'"
     ]
    }
   ],
   "source": [
    "for name, clf in zip(names, classifiers):\n",
    "    print(name)\n",
    "    clf.fit(X_train, y_train)\n",
    "    clf.apply"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "len(clean_data.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [default]",
   "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
}