voting_ensemble.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# voting_ensemble_soft"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from __future__      import division\n",
    "from IPython.display import display\n",
    "from matplotlib      import pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "import numpy  as np\n",
    "import pandas as pd\n",
    "import random, sys, os, re\n",
    "\n",
    "from sklearn.ensemble         import VotingClassifier\n",
    "\n",
    "from sklearn.ensemble         import GradientBoostingClassifier\n",
    "from sklearn.ensemble         import ExtraTreesClassifier\n",
    "from sklearn.linear_model     import LogisticRegression\n",
    "from sklearn.linear_model     import LogisticRegressionCV\n",
    "import xgboost as xgb\n",
    "\n",
    "from sklearn.cross_validation import StratifiedKFold\n",
    "from sklearn.grid_search      import RandomizedSearchCV, GridSearchCV\n",
    "from sklearn.cross_validation import cross_val_predict, permutation_test_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "SEED   = 97\n",
    "scale  = False \n",
    "minmax = False\n",
    "norm   = False\n",
    "nointercept = False\n",
    "engineering = False\n",
    "\n",
    "N_CLASSES = 2\n",
    "\n",
    "submission_filename = \"../submissions/submission_voting_ensemble_softWgtd.csv\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false
   },
   "source": [
    "# Load the training data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from load_blood_data import load_blood_data\n",
    "\n",
    "y_train, X_train = load_blood_data(train=True, SEED   = SEED, \n",
    "                                               scale  = scale,\n",
    "                                               minmax = minmax,\n",
    "                                               norm   = norm,\n",
    "                                               nointercept = nointercept,\n",
    "                                               engineering = engineering)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "StatifiedCV = StratifiedKFold(y            = y_train, \n",
    "                              n_folds      = 10, \n",
    "                              shuffle      = True, \n",
    "                              random_state = SEED)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 164 ms, sys: 20 ms, total: 184 ms\n",
      "Wall time: 401 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "random.seed(SEED)\n",
    "\n",
    "# -------------------------------- estimators ----------------------------------------\n",
    "gbc = GradientBoostingClassifier(loss                     = 'exponential',\n",
    "                                 learning_rate            = 0.15, \n",
    "                                 n_estimators             = 175, \n",
    "                                 max_depth                = 1, \n",
    "                                 subsample                = 0.75, \n",
    "                                 \n",
    "                                 min_samples_split        = 2, \n",
    "                                 min_samples_leaf         = 1, \n",
    "                                 #min_weight_fraction_leaf = 0.0, \n",
    "                                 init                     = None, \n",
    "                                 random_state             = SEED, \n",
    "                                 max_features             = None, \n",
    "                                 verbose                  = 0, \n",
    "                                 max_leaf_nodes           = None, \n",
    "                                 warm_start               = False) \n",
    "                                 #presort                  = 'auto')\n",
    "    \n",
    "etc = ExtraTreesClassifier(n_estimators             = 10, \n",
    "                           criterion                = 'entropy', \n",
    "                           max_depth                = 7, \n",
    "                           bootstrap                = True, \n",
    "                           max_features             = None,\n",
    "                           \n",
    "                           min_samples_split        = 2, \n",
    "                           min_samples_leaf         = 1, \n",
    "                           #min_weight_fraction_leaf = 0.0, \n",
    "                            \n",
    "                           max_leaf_nodes           = None, \n",
    "                           oob_score                = False, \n",
    "                           n_jobs                   = -1, \n",
    "                           random_state             = SEED, \n",
    "                           verbose                  = 0) \n",
    "                           #warm_start               = False, \n",
    "                           #class_weight             = None)\n",
    "        \n",
    "xgbc = xgb.XGBClassifier(learning_rate    = 0.1,\n",
    "                         n_estimators     = 50, \n",
    "                         max_depth        = 5,\n",
    "                         subsample        = 0.25,\n",
    "                         colsample_bytree = 0.75,\n",
    "                         gamma            = 0, \n",
    "                        \n",
    "                         nthread          = 1, \n",
    "                         objective        = 'binary:logistic', \n",
    "                         min_child_weight = 1, \n",
    "                         max_delta_step   = 0, \n",
    "                         base_score       = 0.5, \n",
    "                         seed             = SEED,\n",
    "                         silent           = True,\n",
    "                         missing          = None)\n",
    "\n",
    "logit = LogisticRegression(penalty           = 'l2',          \n",
    "                         dual              = False, \n",
    "                         C                 = 0.001, \n",
    "                         fit_intercept     = True,\n",
    "                         solver            = 'liblinear',   \n",
    "                         max_iter          = 50, \n",
    "                         intercept_scaling = 1,\n",
    "                         \n",
    "                         tol               = 0.0001, \n",
    "                         class_weight      = None, \n",
    "                         random_state      = SEED, \n",
    "                         multi_class       = 'ovr', \n",
    "                         verbose           = 0, \n",
    "                         warm_start        = False, \n",
    "                         n_jobs            = -1)\n",
    "\n",
    "logitCV = LogisticRegressionCV(Cs            = 10, \n",
    "                           cv                = 10,\n",
    "                           \n",
    "                           fit_intercept     = True, \n",
    "                           penalty           = 'l2',\n",
    "                           solver            = 'liblinear', \n",
    "                           max_iter          = 50,\n",
    "                            \n",
    "                           dual              = False, \n",
    "                           scoring           = None, \n",
    "                           tol               = 0.0001, \n",
    "                           class_weight      = None, \n",
    "                           n_jobs            = -1, \n",
    "                           verbose           = 0, \n",
    "                           refit             = True, \n",
    "                           intercept_scaling = 1.0, \n",
    "                           multi_class       = 'ovr', \n",
    "                           random_state      = SEED)\n",
    "\n",
    "# -------------------------------- VotingClassifier ----------------------------------------\n",
    "\n",
    "estimator_list = [('gbc', gbc), ('etc', etc), ('xgbc', xgbc), ('logit', logit), ('logitCV',logitCV)]\n",
    "weights_list   = [      1,          0.75,           0.75,               2,                  1]\n",
    "\n",
    "clf = VotingClassifier(estimators = estimator_list, \n",
    "                       voting     = 'soft', \n",
    "                       weights    = weights_list)\n",
    "\n",
    "clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# from sklearn_utilities import GridSearchHeatmap\n",
    "\n",
    "# GridSearchHeatmap(grid_clf, y_key='learning_rate', x_key='n_estimators')\n",
    "\n",
    "# from sklearn_utilities import plot_validation_curves\n",
    "\n",
    "# plot_validation_curves(grid_clf, param_grid, X_train, y_train, ylim = (0.0, 1.05))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmYAAAGJCAYAAAAg1v9AAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSGEWERER0RMpzCIiIiJ64t/ZFI3x+jv9ywAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f284d890e50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 7.36 s, sys: 1.49 s, total: 8.85 s\n",
      "Wall time: 31 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "try:\n",
    "    from sklearn_utilities import plot_learning_curve\n",
    "except:\n",
    "    import imp, os\n",
    "    util = imp.load_source('sklearn_utilities', os.path.expanduser('~/Dropbox/Python/sklearn_utilities.py'))\n",
    "    from sklearn_utilities import plot_learning_curve\n",
    "\n",
    "plot_learning_curve(estimator   = clf, \n",
    "                    title       = None, \n",
    "                    X           = X_train, \n",
    "                    y           = y_train, \n",
    "                    ylim        = (0.0, 1.10), \n",
    "                    cv          = StratifiedKFold(y            = y_train, \n",
    "                                                  n_folds      = 10, \n",
    "                                                  shuffle      = True, \n",
    "                                                  random_state = SEED), \n",
    "                    train_sizes = np.linspace(.1, 1.0, 5),\n",
    "                    n_jobs      = 1)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training set predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1.64 s, sys: 276 ms, total: 1.92 s\n",
      "Wall time: 5.23 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "train_preds = cross_val_predict(estimator    = clf, \n",
    "                                X            = X_train, \n",
    "                                y            = y_train, \n",
    "                                cv           = StatifiedCV, \n",
    "                                n_jobs       = 1, \n",
    "                                verbose      = 0, \n",
    "                                fit_params   = None, \n",
    "                                pre_dispatch = '2*n_jobs')\n",
    "\n",
    "y_true, y_pred   = y_train, train_preds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[417  21]\n",
      " [102  36]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfQAAAHECAYAAADPrNgnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f284d7898d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy 0.7865, mis-class rate 0.2135\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "cm = confusion_matrix(y_true, y_pred, labels=None)\n",
    "print cm\n",
    "\n",
    "try:\n",
    "    from sklearn_utilities import plot_confusion_matrix\n",
    "except:\n",
    "    import imp, os\n",
    "    util = imp.load_source('sklearn_utilities', os.path.expanduser('~/Dropbox/Python/sklearn_utilities.py'))\n",
    "    from sklearn_utilities import plot_confusion_matrix\n",
    "\n",
    "plot_confusion_matrix(cm, ['Did not Donate','Donated'])\n",
    "\n",
    "accuracy = round(np.trace(cm)/float(np.sum(cm)),4)\n",
    "misclass = 1 - accuracy\n",
    "print(\"Accuracy {}, mis-class rate {}\".format(accuracy,misclass))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmUAAAGJCAYAAADL4URDAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\niQBKykREREQiwP8H8SSbPJ5qvTwAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f284d789f10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import roc_curve\n",
    "from sklearn.metrics import roc_auc_score\n",
    "from sklearn.metrics import log_loss\n",
    "from sklearn.metrics import f1_score\n",
    "\n",
    "fpr, tpr, thresholds = roc_curve(y_true, y_pred, pos_label=None)\n",
    "\n",
    "\n",
    "plt.figure(figsize=(10,6))\n",
    "plt.plot([0, 1], [0, 1], 'k--')\n",
    "plt.plot(fpr, tpr)\n",
    "\n",
    "AUC = roc_auc_score(y_true, y_pred, average='macro')\n",
    "plt.text(x=0.6,y=0.4,s=\"AUC         {:.4f}\"\\\n",
    "         .format(AUC),\n",
    "        fontsize=16)\n",
    "\n",
    "plt.text(x=0.6,y=0.3,s=\"accuracy {:.2f}%\"\\\n",
    "         .format(accuracy*100),\n",
    "        fontsize=16)\n",
    "\n",
    "logloss = log_loss(y_true, y_pred)\n",
    "plt.text(x=0.6,y=0.2,s=\"LogLoss   {:.4f}\"\\\n",
    "         .format(logloss),\n",
    "        fontsize=16)\n",
    "\n",
    "f1 = f1_score(y_true, y_pred)\n",
    "plt.text(x=0.6,y=0.1,s=\"f1             {:.4f}\"\\\n",
    "         .format(f1),\n",
    "        fontsize=16)\n",
    "\n",
    "plt.xlabel('False positive rate')\n",
    "plt.ylabel('True positive rate')\n",
    "plt.title('ROC curve')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 2min 40s, sys: 29.1 s, total: 3min 9s\n",
      "Wall time: 8min 44s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "score, permutation_scores, pvalue = permutation_test_score(estimator      = clf, \n",
    "                                                           X              = X_train.values.astype(np.float32), \n",
    "                                                           y              = y_train, \n",
    "                                                           cv             = StatifiedCV, \n",
    "                                                           labels         = None,\n",
    "                                                           random_state   = SEED,\n",
    "                                                           verbose        = 0,\n",
    "                                                           n_permutations = 100, \n",
    "                                                           scoring        = None,\n",
    "                                                           n_jobs         = 1) \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABIQAAAHuCAYAAADuhdetAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f284d64b610>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20,8))\n",
    "plt.hist(permutation_scores, 20, label='Permutation scores')\n",
    "\n",
    "ylim = plt.ylim()\n",
    "plt.plot(2 * [score], ylim, '--g', linewidth=3,\n",
    "         label='Classification Score (pvalue {:.4f})'.format(pvalue))\n",
    "         \n",
    "plt.plot(2 * [1. / N_CLASSES], ylim, 'r', linewidth=7, label='Luck')\n",
    "\n",
    "plt.ylim(ylim)\n",
    "plt.legend(loc='center',fontsize=16)\n",
    "plt.xlabel('Score')\n",
    "plt.show()\n",
    "\n",
    "# find mean and stdev of the scores\n",
    "from scipy.stats import norm\n",
    "mu, std = norm.fit(permutation_scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "voting_ensemble_softWgtd   ,        ,   0.7865 ,  7.3755 , 0.6065 , 0.3692 , 0.7600 , 0.0013\n"
     ]
    }
   ],
   "source": [
    "# format for scores.csv file\n",
    "import re\n",
    "algo = re.search(r\"submission_(.*?)\\.csv\", submission_filename).group(1)\n",
    "print(\"{: <26} ,        ,   {:.4f} ,  {:.4f} , {:.4f} , {:.4f} , {:.4f} , {:.4f}\"\\\n",
    "      .format(algo,accuracy,logloss,AUC,f1,mu,std))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# --------------------------------------------------------------------------------------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# Test Set Predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Re-fit with the full training set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "VotingClassifier(estimators=[('gbc', GradientBoostingClassifier(init=None, learning_rate=0.15, loss='exponential',\n",
       "              max_depth=1, max_features=None, max_leaf_nodes=None,\n",
       "              min_samples_leaf=1, min_samples_split=2,\n",
       "              min_weight_fraction_leaf=0.0, n_estimators=175,\n",
       "              pres...ate=97,\n",
       "           refit=True, scoring=None, solver='liblinear', tol=0.0001,\n",
       "           verbose=0))],\n",
       "         voting='soft', weights=[1, 0.75, 0.75, 2, 1])"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#clf.set_params(**clf_params)\n",
    "clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load the test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from load_blood_data import load_blood_data\n",
    "\n",
    "X_test, IDs = load_blood_data(train=False, SEED   = SEED, \n",
    "                                           scale  = scale,\n",
    "                                           minmax = minmax,\n",
    "                                           norm   = norm,\n",
    "                                           nointercept = nointercept,\n",
    "                                           engineering = engineering)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predict the test set with the fitted model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 0 0 0 1 1 0 0 0 0]\n",
      "[[ 0.53386636  0.46613364]\n",
      " [ 0.85390385  0.14609615]\n",
      " [ 0.7004128   0.29958721]\n",
      " [ 0.72391782  0.27608219]\n",
      " [ 0.4631666   0.5368334 ]\n",
      " [ 0.22789016  0.77210984]\n",
      " [ 0.63663312  0.36336688]\n",
      " [ 0.87954593  0.12045408]\n",
      " [ 0.98506669  0.01493331]\n",
      " [ 0.95267967  0.04732033]]\n",
      "[0.46613364447000799, 0.14609615014001676, 0.29958720575167985, 0.27608218508709048, 0.53683340216646158, 0.77210984281716322, 0.36336688007603668, 0.12045407518811148, 0.014933309338141213, 0.047320326540905892]\n"
     ]
    }
   ],
   "source": [
    "y_pred = clf.predict(X_test)\n",
    "print(y_pred[:10])\n",
    "\n",
    "try:\n",
    "    y_pred_probs  = clf.predict_proba(X_test)\n",
    "    print(y_pred_probs[:10])\n",
    "    donate_probs  = [prob[1] for prob in y_pred_probs]\n",
    "except Exception,e:\n",
    "    print(e)\n",
    "    donate_probs = [0.65 if x>0 else 1-0.65 for x in y_pred]\n",
    "    \n",
    "print(donate_probs[:10])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Create the submission file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "assert len(IDs)==len(donate_probs)\n",
    "\n",
    "f = open(submission_filename, \"w\")\n",
    "\n",
    "f.write(\",Made Donation in March 2007\\n\")\n",
    "for ID, prob in zip(IDs, donate_probs):\n",
    "    f.write(\"{},{}\\n\".format(ID,prob))\n",
    "    \n",
    "f.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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
}