KMeans.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# KMeans"
   ]
  },
  {
   "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.cluster          import KMeans\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",
    "scale  = False \n",
    "minmax = False\n",
    "norm   = False\n",
    "nointercept = True\n",
    "engineering = False\n",
    "\n",
    "N_CLASSES = 2\n",
    "\n",
    "submission_filename = \"../submissions/submission_KMeans.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 32 ms, sys: 24 ms, total: 56 ms\n",
      "Wall time: 174 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "random.seed(SEED)\n",
    "\n",
    "clf = KMeans(n_clusters           = 2, \n",
    "             \n",
    "             init                 = 'k-means++', # 'k-means++', 'random' or an ndarray\n",
    "             n_init               = 10, \n",
    "             max_iter             = 300, \n",
    "             copy_x               = True,\n",
    "             precompute_distances = 'auto',      # 'auto', True, False\n",
    "             \n",
    "             tol                  = 0.0001,\n",
    "             verbose              = 0, \n",
    "             random_state         = SEED, \n",
    "             n_jobs               = 1)\n",
    "\n",
    "\n",
    "# param_grid = dict(init                 = ['k-means++', 'random'],\n",
    "#                   n_init               = [10, 50],\n",
    "#                   max_iter             = [150, 300, 600],\n",
    "#                   copy_x               = [True, False],\n",
    "#                   precompute_distances = ['auto', True, False])\n",
    "\n",
    "# grid_clf = GridSearchCV(estimator  = clf, \n",
    "#                         param_grid = param_grid,\n",
    "#                         n_jobs     = -1,  \n",
    "#                         cv         = StatifiedCV,\n",
    "#                         verbose    = 0\n",
    "#                        )\n",
    "\n",
    "# grid_clf.fit(X = X_train, y = y_train)\n",
    "\n",
    "# print(\"clf_params = {}\".format(grid_clf.best_params_))\n",
    "# print(\"score: {}\".format(round(grid_clf.best_score_, 4)))\n",
    "# print\n",
    "\n",
    "# clf = grid_clf.best_estimator_\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "clf_params = {'copy_x':   True, \n",
    "              'init':     'k-means++', \n",
    "              'max_iter': 150, \n",
    "              'n_init':   10, \n",
    "              'precompute_distances': 'auto', \n",
    "              'n_jobs':   -1}\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='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": [
    {
     "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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\nYWMmSZJUJ2zMJEmS6sT/D15NF8AOoZBIAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f47bab5a810>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 2.13 s, sys: 972 ms, total: 3.1 s\n",
      "Wall time: 8.87 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 408 ms, sys: 200 ms, total: 608 ms\n",
      "Wall time: 1.75 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": [
      "[[361  77]\n",
      " [102  36]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfsAAAHECAYAAAA+p4OqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nABBO7AEgnNgDQLgDOyy7/q+uwsUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f47b9229f50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy 0.6892, mis-class rate 0.3108\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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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f47b9229210>"
      ]
     },
     "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": "iVBORw0KGgoAAAANSUhEUgAABIEAAAHuCAYAAAAIrBzpAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEAgAAABgAoRAAAAAABPwfzKoT32SFHocAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f47c9933110>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 38.8 s, sys: 21.4 s, total: 1min\n",
      "Wall time: 2min 55s\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": [
      "KMeans                     ,        ,   0.6892 ,  10.7335 , 0.5425 , 0.2869 , -69478760.9634 , 0.0000\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": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KMeans(copy_x=True, init='k-means++', max_iter=150, n_clusters=2, n_init=10,\n",
       "    n_jobs=-1, precompute_distances='auto', random_state=97, tol=0.0001,\n",
       "    verbose=0)"
      ]
     },
     "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": {},
   "source": [
    "## Load the test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1 0 0 1 1 1 0 0 0 0]\n",
      "'KMeans' object has no attribute 'predict_proba'\n",
      "[0.65, 0.35, 0.35, 0.65, 0.65, 0.65, 0.35, 0.35, 0.35, 0.35]\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": 78,
   "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
}