main.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Pre-process"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Load library and define some functions needed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-09T04:37:39.427159Z",
     "start_time": "2017-12-09T04:37:38.566676Z"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import lightgbm as lgb\n",
    "import matplotlib.pyplot as plt\n",
    "import datetime as dt\n",
    "import random\n",
    "\n",
    "import json\n",
    "import os.path\n",
    "import gc\n",
    "\n",
    "from script.lib.FeatureProducer import FeatureProducer \n",
    "MODEL_FILE_NAME = 'model.txt'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-09T04:37:39.507481Z",
     "start_time": "2017-12-09T04:37:39.428873Z"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def custom_cv(params, train_set, hold_out_set=None, k_fold=5, num_boost_round=20):\n",
    "    x_train = train_set.data\n",
    "    y_train = train_set.label\n",
    "    n = x_train.shape[0]\n",
    "    unit = n // k_fold\n",
    "\n",
    "    cv_scores = []\n",
    "    for k in range(k_fold):\n",
    "        ##### !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n",
    "        if k < k_fold - 1:\n",
    "            continue\n",
    "        ##### !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n",
    "        x_cv_valid = None\n",
    "        y_cv_valid = None\n",
    "        if k == k_fold - 1:\n",
    "            x_cv_valid = x_train[unit * k:]\n",
    "            y_cv_valid = y_train[unit * k:]\n",
    "        else:\n",
    "            x_cv_valid = x_train[unit * k: unit * (k + 1)]\n",
    "            y_cv_valid = y_train[unit * k: unit * (k + 1)]\n",
    "\n",
    "        x_cv_train = None\n",
    "        y_cv_train = None\n",
    "        if k == 0:\n",
    "            x_cv_train = x_train[unit * (k + 1):]\n",
    "            y_cv_train = y_train[unit * (k + 1):]\n",
    "        elif k == k_fold - 1:\n",
    "            x_cv_train = x_train[:unit * k]\n",
    "            y_cv_train = y_train[:unit * k]\n",
    "        else:\n",
    "            x_cv_train = x_train[:unit * k].append(x_train[unit * (k + 1):])\n",
    "            y_cv_train = y_train[:unit * k].append(y_train[unit * (k + 1):])\n",
    "\n",
    "        cv_train_set = lgb.Dataset(x_cv_train, y_cv_train)\n",
    "        cv_valid_set = lgb.Dataset(x_cv_valid, y_cv_valid)\n",
    "        watchlist = [cv_valid_set]\n",
    "\n",
    "        # not tested yet\n",
    "        if hold_out_set is not None:\n",
    "            watchlist.append(hold_out_set)\n",
    "        model = lgb.train(params, train_set=cv_train_set, valid_sets=watchlist,\n",
    "                          num_boost_round=num_boost_round, verbose_eval=5)\n",
    "        print(model.best_score)\n",
    "        cv_scores.append(model.best_score['valid_1']['auc'])\n",
    "\n",
    "    tip_txt = '[CV]'\n",
    "    tip_txt += ' ' + str(cv_scores)\n",
    "    mean_cv_score = np.mean(cv_scores)\n",
    "    tip_txt += '{ auc score=' + str(mean_cv_score) + ' }'\n",
    "    print(tip_txt)\n",
    "\n",
    "    return mean_cv_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-09T04:37:39.561212Z",
     "start_time": "2017-12-09T04:37:39.508835Z"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def custom_grid_search(params, own_grid_params, train_set, valid_set, num_boost_round=20):\n",
    "    keys = []\n",
    "    values = [list()]\n",
    "    for key, value in own_grid_params.items():\n",
    "        keys.append(key)\n",
    "        new_values = []\n",
    "        for item in values:\n",
    "            for val in value:\n",
    "                new_values.append(item + [val])\n",
    "        values = new_values\n",
    "\n",
    "    watchlist = [valid_set]\n",
    "    grid_best_params = None\n",
    "    grid_best_score = None\n",
    "\n",
    "    for comb in values:\n",
    "        own_params = {}\n",
    "        for idx in range(len(keys)):\n",
    "            own_params[keys[idx]] = comb[idx]\n",
    "            params[keys[idx]] = comb[idx]\n",
    "\n",
    "        cv_score = custom_cv(params, train_set, valid_set, k_fold=4, num_boost_round=num_boost_round)\n",
    "\n",
    "        tip_txt = '[GridSearch]'\n",
    "        for idx in range(len(keys)):\n",
    "            tip_txt += ' ' + str(keys[idx]) + '=' + str(comb[idx])\n",
    "        tip_txt += ' { best_score: ' + str(cv_score) + ' }'\n",
    "        print(tip_txt)\n",
    "\n",
    "        if grid_best_score is None or cv_score > grid_best_score:\n",
    "            grid_best_params, grid_best_score = own_params, cv_score\n",
    "\n",
    "    tip_txt = '[GS Best Result]'\n",
    "    for key, val in grid_best_params.items():\n",
    "        tip_txt += ' ' + str(key) + '=' + str(val)\n",
    "    tip_txt += ' { best_score: ' + str(grid_best_score) + ' }'\n",
    "    print(tip_txt)\n",
    "    \n",
    "    return grid_best_params"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## init"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-09T04:37:39.622272Z",
     "start_time": "2017-12-09T04:37:39.562603Z"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "fp = FeatureProducer(root='./data')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## load"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-09T04:37:59.853274Z",
     "start_time": "2017-12-09T04:37:39.623587Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "12:37:59 INFO << load raw data in 20.17s\n"
     ]
    }
   ],
   "source": [
    "fp.load_raw()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## preprocess"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-09T04:39:39.514748Z",
     "start_time": "2017-12-09T04:37:59.854834Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "12:38:29 INFO << parse train_df in 30.07s\n",
      "12:38:46 INFO << parse test_df in 16.36s\n",
      "12:38:46 INFO << parse members_df in 0.26s\n",
      "12:39:07 INFO << parse songs_df in 20.12s\n",
      "12:39:10 INFO << parse song_extra_info_df in 3.02s\n"
     ]
    }
   ],
   "source": [
    "fp.pre_process()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## feature engineering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-09T04:40:29.866358Z",
     "start_time": "2017-12-09T04:39:39.516524Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "12:40:07 INFO << parse engineering_df in 27.83s\n",
      "12:40:29 INFO << parse engineering_df in 22.38s\n"
     ]
    }
   ],
   "source": [
    "fp.feature_engineering()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## type to category"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-09T04:40:44.525983Z",
     "start_time": "2017-12-09T04:40:29.868242Z"
    }
   },
   "outputs": [],
   "source": [
    "fp.self_fit_transform()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-09T04:40:46.227114Z",
     "start_time": "2017-12-09T04:40:44.527640Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 7377418 entries, 0 to 7377417\n",
      "Data columns (total 42 columns):\n",
      "msno                        category\n",
      "song_id                     category\n",
      "source_system_tab           category\n",
      "source_screen_name          category\n",
      "source_type                 category\n",
      "target                      int64\n",
      "1h_source                   int64\n",
      "1h_system_tab               int64\n",
      "1h_screen_name              int64\n",
      "1h_source_type              int64\n",
      "song_length                 float64\n",
      "genre_ids                   category\n",
      "artist_name                 category\n",
      "composer                    category\n",
      "lyricist                    category\n",
      "language                    category\n",
      "is_featured                 float64\n",
      "artist_count                float64\n",
      "artist_composer             float64\n",
      "artist_composer_lyricist    float64\n",
      "song_lang_boolean           float64\n",
      "genre_count                 float64\n",
      "composer_count              float64\n",
      "lyricist_count              float64\n",
      "1h_lang                     float64\n",
      "1h_song_length              float64\n",
      "song_year                   float64\n",
      "1h_song_year                float64\n",
      "city                        int64\n",
      "bd                          category\n",
      "gender                      category\n",
      "registered_via              int64\n",
      "expiration_date             category\n",
      "membership_days             int64\n",
      "registration_year           int64\n",
      "registration_month          int64\n",
      "expiration_year             int64\n",
      "expiration_month            int64\n",
      "1h_via                      int64\n",
      "play_count                  int64\n",
      "track_count                 float64\n",
      "cover_lang                  float64\n",
      "dtypes: category(13), float64(15), int64(14)\n",
      "memory usage: 1.9 GB\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 2556790 entries, 0 to 2556789\n",
      "Data columns (total 42 columns):\n",
      "id                          int64\n",
      "msno                        category\n",
      "song_id                     category\n",
      "source_system_tab           category\n",
      "source_screen_name          category\n",
      "source_type                 category\n",
      "1h_source                   int64\n",
      "1h_system_tab               int64\n",
      "1h_screen_name              int64\n",
      "1h_source_type              int64\n",
      "song_length                 float64\n",
      "genre_ids                   category\n",
      "artist_name                 category\n",
      "composer                    category\n",
      "lyricist                    category\n",
      "language                    category\n",
      "is_featured                 float64\n",
      "artist_count                float64\n",
      "artist_composer             float64\n",
      "artist_composer_lyricist    float64\n",
      "song_lang_boolean           float64\n",
      "genre_count                 float64\n",
      "composer_count              float64\n",
      "lyricist_count              float64\n",
      "1h_lang                     float64\n",
      "1h_song_length              float64\n",
      "song_year                   float64\n",
      "1h_song_year                float64\n",
      "city                        int64\n",
      "bd                          category\n",
      "gender                      category\n",
      "registered_via              int64\n",
      "expiration_date             category\n",
      "membership_days             int64\n",
      "registration_year           int64\n",
      "registration_month          int64\n",
      "expiration_year             int64\n",
      "expiration_month            int64\n",
      "1h_via                      int64\n",
      "play_count                  int64\n",
      "track_count                 float64\n",
      "cover_lang                  float64\n",
      "dtypes: category(13), float64(15), int64(14)\n",
      "memory usage: 661.5 MB\n"
     ]
    }
   ],
   "source": [
    "train_df = fp.train_df\n",
    "test_df = fp.test_df\n",
    "\n",
    "assert(train_df.shape[1] == test_df.shape[1])\n",
    "train_df.info()\n",
    "test_df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LGBM training"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Correlation over all features (underconstruction...)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For categorical feature, need to numerize them.\n",
    "Also, for all numeric features, it's necessary to be normalized."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-09T04:41:02.165392Z",
     "start_time": "2017-12-09T04:40:46.228765Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAo8AAAJCCAYAAACh5i0RAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nNv8/LTM7FghDFZIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ff92eb4bcc0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "corr = train_df.corr()\n",
    "\n",
    "f, ax = plt.subplots(figsize=(10, 8))\n",
    "sns.heatmap(corr, mask=np.zeros_like(corr, dtype=np.bool), cmap=sns.diverging_palette(220, 10, as_cmap=True),\n",
    "            square=True, ax=ax)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prepare"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-09T04:41:03.833340Z",
     "start_time": "2017-12-09T04:41:02.166987Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>msno</th>\n",
       "      <th>song_id</th>\n",
       "      <th>source_system_tab</th>\n",
       "      <th>source_screen_name</th>\n",
       "      <th>source_type</th>\n",
       "      <th>target</th>\n",
       "      <th>1h_source</th>\n",
       "      <th>1h_system_tab</th>\n",
       "      <th>1h_screen_name</th>\n",
       "      <th>1h_source_type</th>\n",
       "      <th>...</th>\n",
       "      <th>expiration_date</th>\n",
       "      <th>membership_days</th>\n",
       "      <th>registration_year</th>\n",
       "      <th>registration_month</th>\n",
       "      <th>expiration_year</th>\n",
       "      <th>expiration_month</th>\n",
       "      <th>1h_via</th>\n",
       "      <th>play_count</th>\n",
       "      <th>track_count</th>\n",
       "      <th>cover_lang</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>FGtllVqz18RPiwJj/edr2gV78zirAiY/9SmYvia+kCg=</td>\n",
       "      <td>BBzumQNXUHKdEBOB7mAJuzok+IJA1c2Ryg/yzTF6tik=</td>\n",
       "      <td>explore</td>\n",
       "      <td>Explore</td>\n",
       "      <td>online-playlist</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>2017-10-05</td>\n",
       "      <td>2103</td>\n",
       "      <td>2012</td>\n",
       "      <td>1</td>\n",
       "      <td>2017</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>215</td>\n",
       "      <td>84.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Xumu+NIjS6QYVxDS4/t3SawvJ7viT9hPKXmf0RtLNx8=</td>\n",
       "      <td>bhp/MpSNoqoxOIB+/l8WPqu6jldth4DIpCm3ayXnJqM=</td>\n",
       "      <td>my library</td>\n",
       "      <td>Local playlist more</td>\n",
       "      <td>local-playlist</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2017-09-11</td>\n",
       "      <td>2301</td>\n",
       "      <td>2011</td>\n",
       "      <td>5</td>\n",
       "      <td>2017</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>17080.0</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Xumu+NIjS6QYVxDS4/t3SawvJ7viT9hPKXmf0RtLNx8=</td>\n",
       "      <td>JNWfrrC7zNN7BdMpsISKa4Mw+xVJYNnxXh3/Epw7QgY=</td>\n",
       "      <td>my library</td>\n",
       "      <td>Local playlist more</td>\n",
       "      <td>local-playlist</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2017-09-11</td>\n",
       "      <td>2301</td>\n",
       "      <td>2011</td>\n",
       "      <td>5</td>\n",
       "      <td>2017</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>61.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Xumu+NIjS6QYVxDS4/t3SawvJ7viT9hPKXmf0RtLNx8=</td>\n",
       "      <td>2A87tzfnJTSWqD7gIZHisolhe4DMdzkbd6LzO1KHjNs=</td>\n",
       "      <td>my library</td>\n",
       "      <td>Local playlist more</td>\n",
       "      <td>local-playlist</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2017-09-11</td>\n",
       "      <td>2301</td>\n",
       "      <td>2011</td>\n",
       "      <td>5</td>\n",
       "      <td>2017</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>FGtllVqz18RPiwJj/edr2gV78zirAiY/9SmYvia+kCg=</td>\n",
       "      <td>3qm6XTZ6MOCU11x8FIVbAGH5l5uMkT3/ZalWG1oo2Gc=</td>\n",
       "      <td>explore</td>\n",
       "      <td>Explore</td>\n",
       "      <td>online-playlist</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>2017-10-05</td>\n",
       "      <td>2103</td>\n",
       "      <td>2012</td>\n",
       "      <td>1</td>\n",
       "      <td>2017</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>412</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>FGtllVqz18RPiwJj/edr2gV78zirAiY/9SmYvia+kCg=</td>\n",
       "      <td>3Hg5kugV1S0wzEVLAEfqjIV5UHzb7bCrdBRQlGygLvU=</td>\n",
       "      <td>explore</td>\n",
       "      <td>Explore</td>\n",
       "      <td>online-playlist</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>2017-10-05</td>\n",
       "      <td>2103</td>\n",
       "      <td>2012</td>\n",
       "      <td>1</td>\n",
       "      <td>2017</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>1108</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Xumu+NIjS6QYVxDS4/t3SawvJ7viT9hPKXmf0RtLNx8=</td>\n",
       "      <td>VkILU0H1h3NMmk9MQrXouNudGk5n8Ls5cqRRuBxeTh4=</td>\n",
       "      <td>my library</td>\n",
       "      <td>Local playlist more</td>\n",
       "      <td>local-playlist</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2017-09-11</td>\n",
       "      <td>2301</td>\n",
       "      <td>2011</td>\n",
       "      <td>5</td>\n",
       "      <td>2017</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>3869</td>\n",
       "      <td>43.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>FGtllVqz18RPiwJj/edr2gV78zirAiY/9SmYvia+kCg=</td>\n",
       "      <td>bPIvRTzfHxH5LgHrStll+tYwSQNVV8PySgA3M1PfTgc=</td>\n",
       "      <td>explore</td>\n",
       "      <td>Explore</td>\n",
       "      <td>online-playlist</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>2017-10-05</td>\n",
       "      <td>2103</td>\n",
       "      <td>2012</td>\n",
       "      <td>1</td>\n",
       "      <td>2017</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>287</td>\n",
       "      <td>33.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>uHqAtShXTRXju5GE8ri3ITsVFepPf8jUoCF7ffNOuqE=</td>\n",
       "      <td>/bU6IRSK+YNlNbaTkxo7bhsb2EDLPrnksdX3ggcZNhI=</td>\n",
       "      <td>my library</td>\n",
       "      <td>Local playlist more</td>\n",
       "      <td>local-library</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2018-03-04</td>\n",
       "      <td>2309</td>\n",
       "      <td>2011</td>\n",
       "      <td>11</td>\n",
       "      <td>2018</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>31</td>\n",
       "      <td>155.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>uHqAtShXTRXju5GE8ri3ITsVFepPf8jUoCF7ffNOuqE=</td>\n",
       "      <td>EbI7xoNxI+3QSsiHxL13zBdgHIJOwa3srHd7cDcnJ0g=</td>\n",
       "      <td>my library</td>\n",
       "      <td>Local playlist more</td>\n",
       "      <td>local-library</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2018-03-04</td>\n",
       "      <td>2309</td>\n",
       "      <td>2011</td>\n",
       "      <td>11</td>\n",
       "      <td>2018</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>738</td>\n",
       "      <td>155.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>uHqAtShXTRXju5GE8ri3ITsVFepPf8jUoCF7ffNOuqE=</td>\n",
       "      <td>t0aT90DlS1TGncgnKoL0SvfAWEr3Dl72QBVcokmKfLc=</td>\n",
       "      <td>my library</td>\n",
       "      <td>Local playlist more</td>\n",
       "      <td>local-library</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2018-03-04</td>\n",
       "      <td>2309</td>\n",
       "      <td>2011</td>\n",
       "      <td>11</td>\n",
       "      <td>2018</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>62.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>uHqAtShXTRXju5GE8ri3ITsVFepPf8jUoCF7ffNOuqE=</td>\n",
       "      <td>8FGjC9W+7F8WjheGZPAwX9RH3+nWSO7DzjM6EB6naOI=</td>\n",
       "      <td>my library</td>\n",
       "      <td>Local playlist more</td>\n",
       "      <td>local-library</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2018-03-04</td>\n",
       "      <td>2309</td>\n",
       "      <td>2011</td>\n",
       "      <td>11</td>\n",
       "      <td>2018</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>875</td>\n",
       "      <td>88.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>TJU0Gfvy7FB+r89bWovPKXTjuApTCiv3xg/tt5shR78=</td>\n",
       "      <td>u6/Pb7X4u7KU4gXrBgGqt8RlRrNNFLn03tLAHyxRxwA=</td>\n",
       "      <td>my library</td>\n",
       "      <td>others</td>\n",
       "      <td>top-hits-for-artist</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>2017-09-30</td>\n",
       "      <td>786</td>\n",
       "      <td>2015</td>\n",
       "      <td>8</td>\n",
       "      <td>2017</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>5740</td>\n",
       "      <td>277.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>3g0bC24RD7QUeALY1IF/e8L6X3fHG298ZFIsCEG9T9Y=</td>\n",
       "      <td>TYhx9eqWklddkLQlApQ5MS9jJCO4H3JHdpISZ3kZSRE=</td>\n",
       "      <td>my library</td>\n",
       "      <td>Local playlist more</td>\n",
       "      <td>local-library</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2017-09-22</td>\n",
       "      <td>2405</td>\n",
       "      <td>2011</td>\n",
       "      <td>2</td>\n",
       "      <td>2017</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>3251</td>\n",
       "      <td>341.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>TJU0Gfvy7FB+r89bWovPKXTjuApTCiv3xg/tt5shR78=</td>\n",
       "      <td>IgMar/mVrJQ+ODFPytDf7jwQMmR09+slyZUdVylRFLc=</td>\n",
       "      <td>my library</td>\n",
       "      <td>Local playlist more</td>\n",
       "      <td>local-playlist</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2017-09-30</td>\n",
       "      <td>786</td>\n",
       "      <td>2015</td>\n",
       "      <td>8</td>\n",
       "      <td>2017</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>6830</td>\n",
       "      <td>277.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>15 rows × 42 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                            msno  \\\n",
       "0   FGtllVqz18RPiwJj/edr2gV78zirAiY/9SmYvia+kCg=   \n",
       "1   Xumu+NIjS6QYVxDS4/t3SawvJ7viT9hPKXmf0RtLNx8=   \n",
       "2   Xumu+NIjS6QYVxDS4/t3SawvJ7viT9hPKXmf0RtLNx8=   \n",
       "3   Xumu+NIjS6QYVxDS4/t3SawvJ7viT9hPKXmf0RtLNx8=   \n",
       "4   FGtllVqz18RPiwJj/edr2gV78zirAiY/9SmYvia+kCg=   \n",
       "5   FGtllVqz18RPiwJj/edr2gV78zirAiY/9SmYvia+kCg=   \n",
       "6   Xumu+NIjS6QYVxDS4/t3SawvJ7viT9hPKXmf0RtLNx8=   \n",
       "7   FGtllVqz18RPiwJj/edr2gV78zirAiY/9SmYvia+kCg=   \n",
       "8   uHqAtShXTRXju5GE8ri3ITsVFepPf8jUoCF7ffNOuqE=   \n",
       "9   uHqAtShXTRXju5GE8ri3ITsVFepPf8jUoCF7ffNOuqE=   \n",
       "10  uHqAtShXTRXju5GE8ri3ITsVFepPf8jUoCF7ffNOuqE=   \n",
       "11  uHqAtShXTRXju5GE8ri3ITsVFepPf8jUoCF7ffNOuqE=   \n",
       "12  TJU0Gfvy7FB+r89bWovPKXTjuApTCiv3xg/tt5shR78=   \n",
       "13  3g0bC24RD7QUeALY1IF/e8L6X3fHG298ZFIsCEG9T9Y=   \n",
       "14  TJU0Gfvy7FB+r89bWovPKXTjuApTCiv3xg/tt5shR78=   \n",
       "\n",
       "                                         song_id source_system_tab  \\\n",
       "0   BBzumQNXUHKdEBOB7mAJuzok+IJA1c2Ryg/yzTF6tik=           explore   \n",
       "1   bhp/MpSNoqoxOIB+/l8WPqu6jldth4DIpCm3ayXnJqM=        my library   \n",
       "2   JNWfrrC7zNN7BdMpsISKa4Mw+xVJYNnxXh3/Epw7QgY=        my library   \n",
       "3   2A87tzfnJTSWqD7gIZHisolhe4DMdzkbd6LzO1KHjNs=        my library   \n",
       "4   3qm6XTZ6MOCU11x8FIVbAGH5l5uMkT3/ZalWG1oo2Gc=           explore   \n",
       "5   3Hg5kugV1S0wzEVLAEfqjIV5UHzb7bCrdBRQlGygLvU=           explore   \n",
       "6   VkILU0H1h3NMmk9MQrXouNudGk5n8Ls5cqRRuBxeTh4=        my library   \n",
       "7   bPIvRTzfHxH5LgHrStll+tYwSQNVV8PySgA3M1PfTgc=           explore   \n",
       "8   /bU6IRSK+YNlNbaTkxo7bhsb2EDLPrnksdX3ggcZNhI=        my library   \n",
       "9   EbI7xoNxI+3QSsiHxL13zBdgHIJOwa3srHd7cDcnJ0g=        my library   \n",
       "10  t0aT90DlS1TGncgnKoL0SvfAWEr3Dl72QBVcokmKfLc=        my library   \n",
       "11  8FGjC9W+7F8WjheGZPAwX9RH3+nWSO7DzjM6EB6naOI=        my library   \n",
       "12  u6/Pb7X4u7KU4gXrBgGqt8RlRrNNFLn03tLAHyxRxwA=        my library   \n",
       "13  TYhx9eqWklddkLQlApQ5MS9jJCO4H3JHdpISZ3kZSRE=        my library   \n",
       "14  IgMar/mVrJQ+ODFPytDf7jwQMmR09+slyZUdVylRFLc=        my library   \n",
       "\n",
       "     source_screen_name          source_type  target  1h_source  \\\n",
       "0               Explore      online-playlist       1          0   \n",
       "1   Local playlist more       local-playlist       1          1   \n",
       "2   Local playlist more       local-playlist       1          1   \n",
       "3   Local playlist more       local-playlist       1          1   \n",
       "4               Explore      online-playlist       1          0   \n",
       "5               Explore      online-playlist       1          0   \n",
       "6   Local playlist more       local-playlist       1          1   \n",
       "7               Explore      online-playlist       1          0   \n",
       "8   Local playlist more        local-library       1          1   \n",
       "9   Local playlist more        local-library       1          1   \n",
       "10  Local playlist more        local-library       1          1   \n",
       "11  Local playlist more        local-library       1          1   \n",
       "12               others  top-hits-for-artist       0          0   \n",
       "13  Local playlist more        local-library       1          1   \n",
       "14  Local playlist more       local-playlist       1          1   \n",
       "\n",
       "    1h_system_tab  1h_screen_name  1h_source_type     ...      \\\n",
       "0               0               0               0     ...       \n",
       "1               1               1               1     ...       \n",
       "2               1               1               1     ...       \n",
       "3               1               1               1     ...       \n",
       "4               0               0               0     ...       \n",
       "5               0               0               0     ...       \n",
       "6               1               1               1     ...       \n",
       "7               0               0               0     ...       \n",
       "8               1               1               1     ...       \n",
       "9               1               1               1     ...       \n",
       "10              1               1               1     ...       \n",
       "11              1               1               1     ...       \n",
       "12              1               0               0     ...       \n",
       "13              1               1               1     ...       \n",
       "14              1               1               1     ...       \n",
       "\n",
       "    expiration_date membership_days registration_year registration_month  \\\n",
       "0        2017-10-05            2103              2012                  1   \n",
       "1        2017-09-11            2301              2011                  5   \n",
       "2        2017-09-11            2301              2011                  5   \n",
       "3        2017-09-11            2301              2011                  5   \n",
       "4        2017-10-05            2103              2012                  1   \n",
       "5        2017-10-05            2103              2012                  1   \n",
       "6        2017-09-11            2301              2011                  5   \n",
       "7        2017-10-05            2103              2012                  1   \n",
       "8        2018-03-04            2309              2011                 11   \n",
       "9        2018-03-04            2309              2011                 11   \n",
       "10       2018-03-04            2309              2011                 11   \n",
       "11       2018-03-04            2309              2011                 11   \n",
       "12       2017-09-30             786              2015                  8   \n",
       "13       2017-09-22            2405              2011                  2   \n",
       "14       2017-09-30             786              2015                  8   \n",
       "\n",
       "   expiration_year expiration_month  1h_via  play_count  track_count  \\\n",
       "0             2017               10       1         215         84.0   \n",
       "1             2017                9       1           1      17080.0   \n",
       "2             2017                9       1           4         61.0   \n",
       "3             2017                9       1           1          1.0   \n",
       "4             2017               10       1         412          6.0   \n",
       "5             2017               10       1        1108         15.0   \n",
       "6             2017                9       1        3869         43.0   \n",
       "7             2017               10       1         287         33.0   \n",
       "8             2018                3       1          31        155.0   \n",
       "9             2018                3       1         738        155.0   \n",
       "10            2018                3       1           2         62.0   \n",
       "11            2018                3       1         875         88.0   \n",
       "12            2017                9       1        5740        277.0   \n",
       "13            2017                9       1        3251        341.0   \n",
       "14            2017                9       1        6830        277.0   \n",
       "\n",
       "    cover_lang  \n",
       "0          1.0  \n",
       "1         10.0  \n",
       "2          2.0  \n",
       "3          1.0  \n",
       "4          1.0  \n",
       "5          1.0  \n",
       "6          1.0  \n",
       "7          1.0  \n",
       "8          2.0  \n",
       "9          2.0  \n",
       "10         2.0  \n",
       "11         1.0  \n",
       "12         3.0  \n",
       "13         1.0  \n",
       "14         3.0  \n",
       "\n",
       "[15 rows x 42 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x = train_df.drop(['target'], axis=1)\n",
    "y = train_df['target']\n",
    "\n",
    "# take the last # rows of train_df as valid set where # means number of rows in test_df\n",
    "x_valid = train_df.drop(['target'], axis=1).tail(test_df.shape[0])\n",
    "y_valid = train_df['target'].tail(test_df.shape[0])\n",
    "\n",
    "x_test = test_df.drop(['id'], axis=1)\n",
    "test_ids = test_df['id']\n",
    "\n",
    "train_df.head(15)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Core"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-09T04:41:49.254594Z",
     "start_time": "2017-12-09T04:41:03.834969Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[5]\tvalid_0's auc: 0.655384\tvalid_1's auc: 0.670802\n",
      "defaultdict(<class 'dict'>, {'valid_0': {'auc': 0.6553838494844193}, 'valid_1': {'auc': 0.67080241566192178}})\n",
      "[CV] [0.67080241566192178]{ auc score=0.670802415662 }\n",
      "0.670802415662\n",
      "[5]\tvalid_0's auc: 0.695442\n"
     ]
    }
   ],
   "source": [
    "train_set = lgb.Dataset(x, y)\n",
    "valid_set = lgb.Dataset(x_valid, y_valid, free_raw_data=False)\n",
    "watchlist = [valid_set]\n",
    "\n",
    "params = dict({\n",
    "    'learning_rate': 0.2,\n",
    "    'application': 'binary',\n",
    "    'min_data_in_leaf': 10,\n",
    "#    'max_depth': 10,\n",
    "    'num_leaves': 2 ** 7,\n",
    "    'max_bin': 255,\n",
    "    'verbosity': 0,\n",
    "    'metric': 'auc'\n",
    "})\n",
    "\n",
    "grid_params = {\n",
    "    'learning_rate': [0.1, 0.2],\n",
    "    'max_depth': [8, 10],\n",
    "}\n",
    "\n",
    "# best_grid_params = custom_grid_search(params, grid_params, train_set, hold_out_set, num_boost_round=20)\n",
    "# for key, val in best_grid_params.items():\n",
    "#     params[key] = best_grid_params[key]\n",
    "\n",
    "cv_score = custom_cv(params, train_set, valid_set, k_fold=4, num_boost_round=5)\n",
    "print(cv_score)\n",
    "\n",
    "model = lgb.train(params, train_set=train_set, valid_sets=watchlist, num_boost_round=5, verbose_eval=5)\n",
    "y_test = model.predict(x_test)\n",
    "\n",
    "# When CV, valid_0 means the front 75% training, the last 25% validating\n",
    "#          valid_1 means the front 75% training, the last \"len(test set)\" validating\n",
    "#\n",
    "# When LGBM running, valid_0 means 100% training, the last \"len(test set)\" validating.\n",
    "#\n",
    "# Use CV valid_0 auc score to predict result!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature importances"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-09T04:41:49.669730Z",
     "start_time": "2017-12-09T04:41:49.256317Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlsAAANgCAYAAAAbD62rAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRERERIeSbEVERER0KMlWRERERIf+fysrWKul3VlpAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ff9296a32e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_df = pd.DataFrame({'features': train_df.columns[train_df.columns != 'target'],\n",
    "                        'importances': model.feature_importance()})\n",
    "plot_df = plot_df.sort_values('importances', ascending=False)\n",
    "\n",
    "plt.figure(figsize=(8, 15))\n",
    "sns.barplot(x = plot_df.importances, y = plot_df.features)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Output"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-12-09T04:42:05.999816Z",
     "start_time": "2017-12-09T04:41:49.673482Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 2556790 entries, 0 to 2556789\n",
      "Data columns (total 2 columns):\n",
      "id        int64\n",
      "target    float64\n",
      "dtypes: float64(1), int64(1)\n",
      "memory usage: 58.5 MB\n"
     ]
    }
   ],
   "source": [
    "submission_df = pd.DataFrame()\n",
    "submission_df['id'] = test_ids\n",
    "submission_df['target'] = y_test\n",
    "# string file compression reduces file size\n",
    "submission_df.to_csv('data/submission.csv.gz', compression = 'gzip', index=False, float_format = '%.5f')\n",
    "submission_df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "3.6.3"
  },
  "toc": {
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "toc_cell": false,
   "toc_position": {
    "height": "929px",
    "left": "0px",
    "right": "1122px",
    "top": "107px",
    "width": "294px"
   },
   "toc_section_display": "block",
   "toc_window_display": true
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "oldHeight": 122,
   "position": {
    "height": "40px",
    "left": "1064px",
    "right": "20px",
    "top": "120px",
    "width": "250px"
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "varInspector_section_display": "none",
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}