scikit_neuralnetwork.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# scikit_neuralnetwork"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "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 sknn.mlp                 import Classifier, Layer\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": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "SEED   = 97\n",
    "scale  = True \n",
    "minmax = False\n",
    "norm   = False\n",
    "nointercept = True\n",
    "engineering = True\n",
    "\n",
    "N_CLASSES = 2\n",
    "\n",
    "submission_filename = \"../submissions/submission_scikit_nn.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": {},
   "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": 15,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1.29 s, sys: 4 ms, total: 1.3 s\n",
      "Wall time: 1.29 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "random.seed(SEED)\n",
    "\n",
    "clf_layers = [\n",
    "              Layer(type = 'Rectifier', name         = 'hidden', \n",
    "                                        units        = 200,\n",
    "                                        weight_decay = None,\n",
    "                                        pieces       = None,\n",
    "                                        dropout      = None),\n",
    "    \n",
    "              Layer(type = 'Softmax',   name         = 'output')\n",
    "             ]\n",
    "\n",
    "clf = Classifier(layers            = clf_layers,\n",
    "                 learning_rate     = 0.01,\n",
    "                 learning_rule     = 'momentum',\n",
    "                 learning_momentum = 0.9,\n",
    "                 loss_type         = u'mse',\n",
    "                 mutator           = None,    # data augmentation function\n",
    "                 \n",
    "                 regularize        = None,\n",
    "                 weight_decay      = None,\n",
    "                 dropout_rate      = None,\n",
    "                 \n",
    "                 batch_size        = 25,\n",
    "                 \n",
    "                 valid_size        = 0.1,    \n",
    "                 valid_set         = None,\n",
    "                 \n",
    "                 n_stable          = 10,      # early stopping after ...\n",
    "                 f_stable          = 0.001,   # validation error change threshold\n",
    "                 n_iter            = 100,     # max epochs\n",
    "                 \n",
    "                 random_state      = SEED, \n",
    "                 debug             = False, \n",
    "                 verbose           = True) \n",
    "\n",
    "\n",
    "\n",
    "# param_grid = dict(learning_rate = [0.0001, 0.001, 0.003, 0.01],\n",
    "#                   learning_rule = ['momentum', 'nesterov'],\n",
    "#                   batch_size    = [1, 10, 100],\n",
    "                  \n",
    "#                   hidden__units = [10, 100, 500, 750],\n",
    "#                   hidden__type  = ['Rectifier', 'Sigmoid', 'Tanh'])\n",
    "\n",
    "# grid_clf = GridSearchCV(estimator  = clf, \n",
    "#                         param_grid = param_grid,\n",
    "#                         n_jobs     = -1,  \n",
    "#                         cv         = StatifiedCV)\n",
    "\n",
    "# grid_clf.fit(X_train.values.astype(np.float32), 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 =  {'learning_rate': 0.01, \n",
    "               'learning_rule': 'nesterov', \n",
    "               'hidden__type': 'Rectifier', \n",
    "               'batch_size': 10, \n",
    "               'hidden__units': 100,\n",
    "               'valid_size': None,\n",
    "               'valid_set': None}\n",
    "clf.set_params(**clf_params)\n",
    "clf.fit(X_train.values.astype(np.float32), y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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": 16,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmYAAAGJCAYAAAAg1v9AAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIiIiIkbE/wBm9RcXxjjGIAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4be860a750>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "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       = 'Learning Curves', \n",
    "                    X           = X_train.values.astype(np.float32), \n",
    "                    y           = y_train, \n",
    "                    ylim        = (0.0, 1.10), \n",
    "                    cv          = StatifiedCV, \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": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 172 ms, sys: 56 ms, total: 228 ms\n",
      "Wall time: 3.71 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "train_preds = cross_val_predict(estimator    = clf, \n",
    "                                X            = X_train.values.astype(np.float32), \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": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[396  42]\n",
      " [ 75  63]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfsAAAHECAYAAAA+p4OqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nFPuIiIiOS7GPiIjouBT7iIiIjvsBwcFp0238RX4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4be97ace90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy 0.7969, mis-class rate 0.2031\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": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmUAAAGJCAYAAADL4URDAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRCKAkjIRERGRCPB/cnCwsLEkbyQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4be832ff50>"
      ]
     },
     "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": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABIMAAAH6CAYAAACUHoxZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4bf30f3e10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 10.1 s, sys: 136 ms, total: 10.2 s\n",
      "Wall time: 4min 6s\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",
    "# find mean and stdev of the scores\n",
    "from scipy.stats import norm\n",
    "mu, std = norm.fit(permutation_scores)\n",
    "\n",
    "\n",
    "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.title('mu={:.4f}, std={:.4f}'.format(mu,std), fontsize=20)\n",
    "plt.legend(loc='center',fontsize=16)\n",
    "plt.xlabel('Score')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "scikit_nn                  ,        ,   0.7969 ,  7.0157 , 0.6803 , 0.5185 , 0.7387 , 0.0116\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": [
    "# Predict the leaderboard score with linear regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "clear all but functions rm(list = setdiff(ls(), lsf.str()))\n",
      "R functions in Dropbox:\n",
      " [1] \"confusion.matrix_ROC.R\" \"eda.R\"                  \"knitr_settings.R\"      \n",
      " [4] \"lm_utilities.R\"         \"lsp.R\"                  \"matrix.power.R\"        \n",
      " [7] \"metrics.R\"              \"multiplot.R\"            \"my.special.functions.R\"\n",
      "[10] \"NOT_IN.R\"               \"SDAFE.R\"                \"signOnToTwitter.R\"     \n",
      "[13] \"source.dropbox.R\"       \"stem_plot.R\"           \n",
      "to load these functions: source(\"/home/george/Dropbox/R_functions/*.R\")\n",
      "\n",
      "To save graphical parameters: opar = par(no.readonly=TRUE)\n",
      "to restore graphical parameters: par(opar)\n"
     ]
    }
   ],
   "source": [
    "# load the R extension\n",
    "%load_ext rpy2.ipython\n",
    "\n",
    "# see http://ipython.readthedocs.org/en/stable/config/extensions/index.html?highlight=rmagic\n",
    "# see http://rpy.sourceforge.net/rpy2/doc-2.4/html/interactive.html#module-rpy2.ipython.rmagic"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Import python variables into R\n",
    "%R -i accuracy,logloss,AUC,f1,mu,std"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "     fit    lwr    upr\n",
       "1 0.8708 -0.037 1.7786\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "# read in the scores.csv file and perform a linear regression with it using this process's variables\n",
    "\n",
    "score_data = read.csv('../input/scores.csv')\n",
    "\n",
    "lm.fit     = lm(leaderboard_score ~ accuracy + logloss + AUC + f1 + mu + std, \n",
    "                data      = score_data, \n",
    "                na.action = na.omit)\n",
    "\n",
    "slm.fit = step(lm.fit, direction = \"both\", trace=0)\n",
    "\n",
    "predicted_leaderboard_score = predict(object   = slm.fit, \n",
    "                                      newdata  = data.frame(accuracy,logloss,AUC,f1,mu,std),\n",
    "                                      interval = \"prediction\", level = 0.99)\n",
    "\n",
    "print(round(predicted_leaderboard_score,4))"
   ]
  },
  {
   "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": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Classifier(batch_size=10, debug=False, dropout_rate=None, f_stable=0.001,\n",
       "      hidden=<sknn.nn.Layer `Rectifier`: name='hidden', units=100>,\n",
       "      layers=[<sknn.nn.Layer `Rectifier`: name='hidden', units=100>, <sknn.nn.Layer `Softmax`: name='output', units=2>],\n",
       "      learning_momentum=0.9, learning_rate=0.01, learning_rule='nesterov',\n",
       "      loss_type=u'mse', mutator=None, n_iter=100, n_stable=10,\n",
       "      output=<sknn.nn.Layer `Softmax`: name='output', units=2>,\n",
       "      random_state=97, regularize=None, valid_set=None, valid_size=None,\n",
       "      verbose=True, weight_decay=None)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.set_params(**clf_params)\n",
    "clf.fit(X_train.values.astype(np.float32), y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load the test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "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": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1 0 0 0 1 1 0 0 0 0]\n",
      "[[  3.69503021e-01   6.30496979e-01]\n",
      " [  9.12029505e-01   8.79704878e-02]\n",
      " [  9.53740537e-01   4.62594479e-02]\n",
      " [  8.28116298e-01   1.71883717e-01]\n",
      " [  3.24051440e-01   6.75948560e-01]\n",
      " [  6.41991720e-02   9.35800850e-01]\n",
      " [  8.63493204e-01   1.36506796e-01]\n",
      " [  9.82012033e-01   1.79879777e-02]\n",
      " [  9.99999881e-01   1.24267245e-07]\n",
      " [  9.88091528e-01   1.19084781e-02]]\n",
      "[0.63049698, 0.087970488, 0.046259448, 0.17188372, 0.67594856, 0.93580085, 0.1365068, 0.017987978, 1.2426725e-07, 0.011908478]\n"
     ]
    }
   ],
   "source": [
    "y_pred = clf.predict(X_test.values.astype(np.float32)).ravel()\n",
    "print(y_pred[:10])\n",
    "\n",
    "try:\n",
    "    y_pred_probs  = clf.predict_proba(X_test.values.astype(np.float32))\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": 31,
   "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
}