Base_Line.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Base Line"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "from scipy import interp\n",
    "\n",
    "from sklearn.externals import joblib\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import roc_curve, auc\n",
    "from IPython.display import display\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "from utils import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.set_option('display.max_columns', 500)\n",
    "pd.set_option('display.max_rows', 500)\n",
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(307511, 317)\n"
     ]
    }
   ],
   "source": [
    "train = joblib.load('models/train.joblib')\n",
    "print(train.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(48744, 316)\n"
     ]
    }
   ],
   "source": [
    "test = joblib.load('models/test.joblib')\n",
    "print(test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "targets = train['TARGET']\n",
    "train_ids = train['SK_ID_CURR']\n",
    "train.drop(columns=['SK_ID_CURR', 'TARGET'], inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_ids = test['SK_ID_CURR']\n",
    "test = test.drop(columns=['SK_ID_CURR'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Drop redundand columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(307511, 295)\n",
      "(48744, 295)\n"
     ]
    }
   ],
   "source": [
    "cols_drop = appartment_mode_cols + appartment_medi_cols\n",
    "train.drop(columns=cols_drop, inplace=True)\n",
    "test.drop(columns=cols_drop, inplace=True)\n",
    "print(train.shape)\n",
    "print(test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fb3af062518>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fb3aef707f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "features = np.array(train)\n",
    "test_features = np.array(test)\n",
    "\n",
    "k_fold = StratifiedKFold(n_splits=3, shuffle=True, random_state=42)\n",
    "\n",
    "i = 0\n",
    "test_predictions = np.zeros(test_features.shape[0])\n",
    "\n",
    "for train_indices, valid_indices in k_fold.split(features, targets):\n",
    "    # Training data for the fold\n",
    "    train_features, train_labels = features[train_indices], targets[train_indices]\n",
    "    # Validation data for the fold\n",
    "    valid_features, valid_labels = features[valid_indices], targets[valid_indices]\n",
    "    \n",
    "    model = LogisticRegression(C=0.001, class_weight='balanced', random_state=4242)\n",
    "    probas_ = model.fit(train_features, train_labels).predict_proba(valid_features)\n",
    "    test_predictions += model.predict_proba(test_features)[:, 1] / k_fold.n_splits\n",
    "\n",
    "    fpr, tpr, thresholds = roc_curve(valid_labels, probas_[:, 1])\n",
    "    roc_auc = auc(fpr, tpr)\n",
    "    plt.plot(fpr, tpr, lw=1, alpha=0.3, label='ROC fold %d (AUC = %0.2f)' % (i, roc_auc))\n",
    "    i += 1\n",
    "\n",
    "\n",
    "plt.plot([0, 1], [0, 1], linestyle='--', lw=2, color='r', label='Luck', alpha=.8)\n",
    "\n",
    "plt.xlim([-0.05, 1.05])\n",
    "plt.ylim([-0.05, 1.05])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Receiver operating characteristic')\n",
    "plt.legend(loc=\"lower right\")\n",
    "plt.figure(figsize=(7,5))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'kaggle competitions submit -c home-credit-default-risk -f /home/inc/work/edu/ndml/capstone/submissions/submission-LogisticRegression.csv -m \"LogisticRegression\"'"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "submission = pd.DataFrame({'SK_ID_CURR': test_ids, 'TARGET': test_predictions})\n",
    "save_prediction(submission, 'LogisticRegression')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}