boosted_logit.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# boosted_logit"
   ]
  },
  {
   "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 AdaBoostClassifier\n",
    "\n",
    "from sklearn.linear_model     import LogisticRegression\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": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "SEED   = 97\n",
    "\n",
    "scale  = False \n",
    "minmax = False\n",
    "norm   = False\n",
    "nointercept = False\n",
    "engineering = True\n",
    "\n",
    "N_CLASSES = 2\n",
    "\n",
    "submission_filename = \"../submissions/submission_boosted_logit_newton.csv\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "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": {
    "collapsed": false
   },
   "source": [
    "# Fit 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
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AdaBoostClassifier(algorithm='SAMME',\n",
       "          base_estimator=LogisticRegression(C=0.001, class_weight=None, dual=False, fit_intercept=True,\n",
       "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
       "          penalty='l2', random_state=97, solver='newton-cg', tol=0.0001,\n",
       "          verbose=0, warm_start=False),\n",
       "          learning_rate=0.1, n_estimators=75, random_state=97)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#%%time\n",
    "\n",
    "random.seed(SEED)\n",
    " \n",
    "# ---------------------------------- base estimator ------------------------------------------------\n",
    "base = LogisticRegression(penalty           = 'l2',          \n",
    "                         dual              = False, \n",
    "                         C                 = 0.001, \n",
    "                         fit_intercept     = True,\n",
    "                         solver            = 'newton-cg',   \n",
    "                         max_iter          = 100, \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",
    "# ---------------------------------- AdaBoostClassifier ------------------------------------------------\n",
    "clf = AdaBoostClassifier(base_estimator = base, \n",
    "                         n_estimators   = 50, \n",
    "                         learning_rate  = 1.0, \n",
    "                         algorithm      = 'SAMME.R', \n",
    "                         random_state   = SEED)\n",
    "    \n",
    "\n",
    "# param_grid = dict(n_estimators  = [25, 50, 75, 100],\n",
    "#                   learning_rate = [0.001, 0.01, 0.1, 1.0],\n",
    "#                   algorithm     = ['SAMME','SAMME.R'])\n",
    "#                   #base__solver  = ['newton-cg', 'lbfgs' 'sag']) \n",
    "\n",
    "# grid_clf = GridSearchCV(estimator  = clf, \n",
    "#                         param_grid = param_grid,\n",
    "#                         n_jobs     = -1,  \n",
    "#                         cv         = StatifiedCV).fit(X_train, y_train)\n",
    "\n",
    "# print(\"clf_params = {}\".format(grid_clf.best_params_))\n",
    "# print(\"score: {}\".format(grid_clf.best_score_))\n",
    "\n",
    "# clf = grid_clf.best_estimator_\n",
    "\n",
    "\n",
    "\n",
    "clf_params = {'n_estimators': 75, 'learning_rate': 0.1, 'algorithm': 'SAMME'}\n",
    "clf.set_params(**clf_params)\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='C', x_key='gamma')\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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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fde69a93910>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 296 ms, sys: 28 ms, total: 324 ms\n",
      "Wall time: 5.29 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",
    "\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 68 ms, sys: 24 ms, total: 92 ms\n",
      "Wall time: 1.56 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": [
      "[[415  23]\n",
      " [109  29]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfsAAAHECAYAAAA+p4OqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nmZnVnIO9mZlZzTnYm5mZ1dwvdKKS1swnIKIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fde9c4ec9d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy 0.7708, mis-class rate 0.2292\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",
    "\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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\nkjIRERGRGKCkTERERCQG/B8J1mAiWPbDEQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fde9c4ecbd0>"
      ]
     },
     "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": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABIQAAAHuCAYAAADuhdetAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nMApCAAAAAAPz/8iCyZl2JPKEAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fde69fa4d90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 5.76 s, sys: 44 ms, total: 5.8 s\n",
      "Wall time: 2min 1s\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",
    "plt.figure(figsize=(20,8))\n",
    "plt.hist(permutation_scores, 20, label='Permutation scores')\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": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "boosted_logit_newton       ,        ,   0.7708 ,  7.9152 , 0.5788 , 0.3053 , 0.7589 , 0.0020\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": {
    "collapsed": true
   },
   "source": [
    "# ----------------------------------------------------------------------------------------\n",
    "\n",
    "# Test Set Predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Re-fit with the full training set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AdaBoostClassifier(algorithm='SAMME',\n",
       "          base_estimator=LogisticRegression(C=0.001, class_weight=None, dual=False, fit_intercept=True,\n",
       "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
       "          penalty='l2', random_state=97, solver='lbfgs', tol=0.0001,\n",
       "          verbose=0, warm_start=False),\n",
       "          learning_rate=0.1, n_estimators=75, random_state=97)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.set_params(**clf_params)\n",
    "clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Load the test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "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": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1 0 0 0 1 1 0 0 0 0]\n",
      "[[ 0.48735838  0.51264162]\n",
      " [ 0.6308336   0.3691664 ]\n",
      " [ 0.54822552  0.45177448]\n",
      " [ 0.55355365  0.44644635]\n",
      " [ 0.49286856  0.50713144]\n",
      " [ 0.42972522  0.57027478]\n",
      " [ 0.5417354   0.4582646 ]\n",
      " [ 0.57710302  0.42289698]\n",
      " [ 0.70756789  0.29243211]\n",
      " [ 0.65128106  0.34871894]]\n",
      "[0.51264161984099843, 0.36916640124703048, 0.45177447923016573, 0.44644634839138397, 0.50713143702095154, 0.5702747820578683, 0.45826459969047345, 0.42289697584494379, 0.29243211009081038, 0.34871894113387852]\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": 104,
   "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": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "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
}