SVC.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SVC"
   ]
  },
  {
   "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.svm              import SVC\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_SVC.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": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 44 ms, sys: 0 ns, total: 44 ms\n",
      "Wall time: 45.2 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "random.seed(SEED)\n",
    " \n",
    "clf = SVC(C                       = 1.0, \n",
    "          kernel                  = 'rbf', \n",
    "          degree                  = 3, \n",
    "          gamma                   = 'auto', \n",
    "          coef0                   = 0.0, \n",
    "          \n",
    "          shrinking               = True, \n",
    "          probability             = True, \n",
    "          tol                     = 0.001, \n",
    "          cache_size              = 2000, \n",
    "          class_weight            = None, \n",
    "          verbose                 = False, \n",
    "          max_iter                = -1, \n",
    "          #decision_function_shape = None, \n",
    "          random_state            = SEED)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# param_grid = dict(gamma  = [0.00001, 0.0001, 0.001, 0.01, 0.1,'auto'],\n",
    "#                   C      = [0.001, 0.01, 0.1, 1.0])\n",
    "\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",
    "\n",
    "# clf_params = {'C': 1.0, 'gamma': 0.0001}\n",
    "# clf.set_params(**clf_params)\n",
    "clf.fit(X_train, y_train)\n",
    "clf_params = clf.get_params()"
   ]
  },
  {
   "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": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/george/.local/lib/python2.7/site-packages/matplotlib/collections.py:590: FutureWarning: elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison\n",
      "  if self._edgecolors == str('face'):\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmYAAAGJCAYAAAAg1v9AAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\njvgNxKeHECb8oW8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f13153166d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 228 ms, sys: 24 ms, total: 252 ms\n",
      "Wall time: 784 ms\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 60 ms, sys: 28 ms, total: 88 ms\n",
      "Wall time: 282 ms\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": [
      "[[423  15]\n",
      " [126  12]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfQAAAHECAYAAADPrNgnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f13132e1490>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy 0.7552, mis-class rate 0.2448\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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\nRERCgIIyERERkRCgoExEREQkBPw/nL2ORZmsr5UAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f13132e1ad0>"
      ]
     },
     "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": "iVBORw0KGgoAAAANSUhEUgAABIMAAAHuCAYAAAAMWczUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f13493fa1d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 788 ms, sys: 36 ms, total: 824 ms\n",
      "Wall time: 9.43 s\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": [
      "SVC                        ,        ,   0.7552 ,  8.4548 , 0.5264 , 0.1455 , 0.7420 , 0.0079\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": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=1.0, cache_size=2000, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape=None, degree=3, gamma='auto', kernel='rbf',\n",
       "  max_iter=-1, probability=True, random_state=97, shrinking=True,\n",
       "  tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 15,
     "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": 16,
   "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": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 0 0 0 0 0 0 0 0 0]\n",
      "[[ 0.75609895  0.24390105]\n",
      " [ 0.75609907  0.24390093]\n",
      " [ 0.79855507  0.20144493]\n",
      " [ 0.75609907  0.24390093]\n",
      " [ 0.75609907  0.24390093]\n",
      " [ 0.75609907  0.24390093]\n",
      " [ 0.7985445   0.2014555 ]\n",
      " [ 0.79852712  0.20147288]\n",
      " [ 0.75609907  0.24390093]\n",
      " [ 0.75609907  0.24390093]]\n",
      "[0.24390105043901128, 0.24390093102993365, 0.20144493209079276, 0.24390093102993365, 0.24390093102993365, 0.24390093102993365, 0.20145549726286036, 0.20147287766194713, 0.24390093102993365, 0.24390093102993365]\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": 13,
   "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
}