5. Голодающие дети Африки.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:43:44.818732",
     "start_time": "2017-07-25T11:43:44.810735"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "pd.options.display.max_columns = 200\n",
    "\n",
    "import numpy as np\n",
    "import pickle\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "import seaborn as sns\n",
    "sns.set()\n",
    "\n",
    "# увеличим дефолтный размер графиков\n",
    "from pylab import rcParams\n",
    "rcParams['figure.figsize'] = 18, 6\n",
    "rcParams['font.size'] = 16\n",
    "rcParams['axes.labelsize'] = 14\n",
    "rcParams['xtick.labelsize'] = 13\n",
    "rcParams['ytick.labelsize'] = 13"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Данные"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Зайдите на https://www.drivendata.org/ и зарегистрируйтесь. Для сегодняшней домашки будем данные брать именно отсюда.\n",
    "\n",
    "Нас интересует конкурс https://www.drivendata.org/competitions/7/pump-it-up-data-mining-the-water-table/page/23/ .\n",
    "В нем представлены данные, собранные [Taarifa](taarifa.org) и [Танзанийским Министерством Воды и Ирригации](https://www.maji.go.tz/?q=en).\n",
    "\n",
    "Постановка задачи следующая: \n",
    "На территории Танзании установлено множество водяных насосов, которые спасают местное население от жажды. В зависимости от того, кем и когда установлен насос, а также зная, как им распоряжаются, можно попытаться предположить, какие из них функционируют, какие нуждаются в ремонте и какие не работают вовсе. \n",
    "\n",
    "Этим мы и займемся, а заодно и прокачаемся в подборе гиперпараметров алгоритмов."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:44:47.787521",
     "start_time": "2017-07-25T11:44:46.914095"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "train_X, train_y = pd.read_csv( # путь к вашему файлу train.csv\n",
    "    'data/WaterTable/train.csv'\n",
    "), pd.read_csv( # путь к вашему файлу trainLabels.csv\n",
    "    'data/WaterTable/trainLabels.csv'\n",
    ")\n",
    "df = pd.merge(train_X, train_y, how='left')\n",
    "\n",
    "df_test = pd.read_csv( # путь к вашему файлу test.csv\n",
    "    'data/WaterTable/test.csv'\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "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>id</th>\n",
       "      <th>amount_tsh</th>\n",
       "      <th>date_recorded</th>\n",
       "      <th>funder</th>\n",
       "      <th>gps_height</th>\n",
       "      <th>installer</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>wpt_name</th>\n",
       "      <th>num_private</th>\n",
       "      <th>basin</th>\n",
       "      <th>subvillage</th>\n",
       "      <th>region</th>\n",
       "      <th>region_code</th>\n",
       "      <th>district_code</th>\n",
       "      <th>lga</th>\n",
       "      <th>ward</th>\n",
       "      <th>population</th>\n",
       "      <th>public_meeting</th>\n",
       "      <th>recorded_by</th>\n",
       "      <th>scheme_management</th>\n",
       "      <th>scheme_name</th>\n",
       "      <th>permit</th>\n",
       "      <th>construction_year</th>\n",
       "      <th>extraction_type</th>\n",
       "      <th>extraction_type_group</th>\n",
       "      <th>extraction_type_class</th>\n",
       "      <th>management</th>\n",
       "      <th>management_group</th>\n",
       "      <th>payment</th>\n",
       "      <th>payment_type</th>\n",
       "      <th>water_quality</th>\n",
       "      <th>quality_group</th>\n",
       "      <th>quantity</th>\n",
       "      <th>quantity_group</th>\n",
       "      <th>source</th>\n",
       "      <th>source_type</th>\n",
       "      <th>source_class</th>\n",
       "      <th>waterpoint_type</th>\n",
       "      <th>waterpoint_type_group</th>\n",
       "      <th>status_group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>69572</td>\n",
       "      <td>6000.0</td>\n",
       "      <td>2011-03-14</td>\n",
       "      <td>Roman</td>\n",
       "      <td>1390</td>\n",
       "      <td>Roman</td>\n",
       "      <td>34.938093</td>\n",
       "      <td>-9.856322</td>\n",
       "      <td>none</td>\n",
       "      <td>0</td>\n",
       "      <td>Lake Nyasa</td>\n",
       "      <td>Mnyusi B</td>\n",
       "      <td>Iringa</td>\n",
       "      <td>11</td>\n",
       "      <td>5</td>\n",
       "      <td>Ludewa</td>\n",
       "      <td>Mundindi</td>\n",
       "      <td>109</td>\n",
       "      <td>True</td>\n",
       "      <td>GeoData Consultants Ltd</td>\n",
       "      <td>VWC</td>\n",
       "      <td>Roman</td>\n",
       "      <td>False</td>\n",
       "      <td>1999</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>vwc</td>\n",
       "      <td>user-group</td>\n",
       "      <td>pay annually</td>\n",
       "      <td>annually</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>enough</td>\n",
       "      <td>spring</td>\n",
       "      <td>spring</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8776</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2013-03-06</td>\n",
       "      <td>Grumeti</td>\n",
       "      <td>1399</td>\n",
       "      <td>GRUMETI</td>\n",
       "      <td>34.698766</td>\n",
       "      <td>-2.147466</td>\n",
       "      <td>Zahanati</td>\n",
       "      <td>0</td>\n",
       "      <td>Lake Victoria</td>\n",
       "      <td>Nyamara</td>\n",
       "      <td>Mara</td>\n",
       "      <td>20</td>\n",
       "      <td>2</td>\n",
       "      <td>Serengeti</td>\n",
       "      <td>Natta</td>\n",
       "      <td>280</td>\n",
       "      <td>NaN</td>\n",
       "      <td>GeoData Consultants Ltd</td>\n",
       "      <td>Other</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "      <td>2010</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>wug</td>\n",
       "      <td>user-group</td>\n",
       "      <td>never pay</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>insufficient</td>\n",
       "      <td>insufficient</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>34310</td>\n",
       "      <td>25.0</td>\n",
       "      <td>2013-02-25</td>\n",
       "      <td>Lottery Club</td>\n",
       "      <td>686</td>\n",
       "      <td>World vision</td>\n",
       "      <td>37.460664</td>\n",
       "      <td>-3.821329</td>\n",
       "      <td>Kwa Mahundi</td>\n",
       "      <td>0</td>\n",
       "      <td>Pangani</td>\n",
       "      <td>Majengo</td>\n",
       "      <td>Manyara</td>\n",
       "      <td>21</td>\n",
       "      <td>4</td>\n",
       "      <td>Simanjiro</td>\n",
       "      <td>Ngorika</td>\n",
       "      <td>250</td>\n",
       "      <td>True</td>\n",
       "      <td>GeoData Consultants Ltd</td>\n",
       "      <td>VWC</td>\n",
       "      <td>Nyumba ya mungu pipe scheme</td>\n",
       "      <td>True</td>\n",
       "      <td>2009</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>vwc</td>\n",
       "      <td>user-group</td>\n",
       "      <td>pay per bucket</td>\n",
       "      <td>per bucket</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>enough</td>\n",
       "      <td>dam</td>\n",
       "      <td>dam</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe multiple</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>67743</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2013-01-28</td>\n",
       "      <td>Unicef</td>\n",
       "      <td>263</td>\n",
       "      <td>UNICEF</td>\n",
       "      <td>38.486161</td>\n",
       "      <td>-11.155298</td>\n",
       "      <td>Zahanati Ya Nanyumbu</td>\n",
       "      <td>0</td>\n",
       "      <td>Ruvuma / Southern Coast</td>\n",
       "      <td>Mahakamani</td>\n",
       "      <td>Mtwara</td>\n",
       "      <td>90</td>\n",
       "      <td>63</td>\n",
       "      <td>Nanyumbu</td>\n",
       "      <td>Nanyumbu</td>\n",
       "      <td>58</td>\n",
       "      <td>True</td>\n",
       "      <td>GeoData Consultants Ltd</td>\n",
       "      <td>VWC</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "      <td>1986</td>\n",
       "      <td>submersible</td>\n",
       "      <td>submersible</td>\n",
       "      <td>submersible</td>\n",
       "      <td>vwc</td>\n",
       "      <td>user-group</td>\n",
       "      <td>never pay</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>dry</td>\n",
       "      <td>dry</td>\n",
       "      <td>machine dbh</td>\n",
       "      <td>borehole</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe multiple</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>non functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>19728</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2011-07-13</td>\n",
       "      <td>Action In A</td>\n",
       "      <td>0</td>\n",
       "      <td>Artisan</td>\n",
       "      <td>31.130847</td>\n",
       "      <td>-1.825359</td>\n",
       "      <td>Shuleni</td>\n",
       "      <td>0</td>\n",
       "      <td>Lake Victoria</td>\n",
       "      <td>Kyanyamisa</td>\n",
       "      <td>Kagera</td>\n",
       "      <td>18</td>\n",
       "      <td>1</td>\n",
       "      <td>Karagwe</td>\n",
       "      <td>Nyakasimbi</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>GeoData Consultants Ltd</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>other</td>\n",
       "      <td>other</td>\n",
       "      <td>never pay</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>seasonal</td>\n",
       "      <td>seasonal</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id  amount_tsh date_recorded        funder  gps_height     installer  \\\n",
       "0  69572      6000.0    2011-03-14         Roman        1390         Roman   \n",
       "1   8776         0.0    2013-03-06       Grumeti        1399       GRUMETI   \n",
       "2  34310        25.0    2013-02-25  Lottery Club         686  World vision   \n",
       "3  67743         0.0    2013-01-28        Unicef         263        UNICEF   \n",
       "4  19728         0.0    2011-07-13   Action In A           0       Artisan   \n",
       "\n",
       "   longitude   latitude              wpt_name  num_private  \\\n",
       "0  34.938093  -9.856322                  none            0   \n",
       "1  34.698766  -2.147466              Zahanati            0   \n",
       "2  37.460664  -3.821329           Kwa Mahundi            0   \n",
       "3  38.486161 -11.155298  Zahanati Ya Nanyumbu            0   \n",
       "4  31.130847  -1.825359               Shuleni            0   \n",
       "\n",
       "                     basin  subvillage   region  region_code  district_code  \\\n",
       "0               Lake Nyasa    Mnyusi B   Iringa           11              5   \n",
       "1            Lake Victoria     Nyamara     Mara           20              2   \n",
       "2                  Pangani     Majengo  Manyara           21              4   \n",
       "3  Ruvuma / Southern Coast  Mahakamani   Mtwara           90             63   \n",
       "4            Lake Victoria  Kyanyamisa   Kagera           18              1   \n",
       "\n",
       "         lga        ward  population public_meeting              recorded_by  \\\n",
       "0     Ludewa    Mundindi         109           True  GeoData Consultants Ltd   \n",
       "1  Serengeti       Natta         280            NaN  GeoData Consultants Ltd   \n",
       "2  Simanjiro     Ngorika         250           True  GeoData Consultants Ltd   \n",
       "3   Nanyumbu    Nanyumbu          58           True  GeoData Consultants Ltd   \n",
       "4    Karagwe  Nyakasimbi           0           True  GeoData Consultants Ltd   \n",
       "\n",
       "  scheme_management                  scheme_name permit  construction_year  \\\n",
       "0               VWC                        Roman  False               1999   \n",
       "1             Other                          NaN   True               2010   \n",
       "2               VWC  Nyumba ya mungu pipe scheme   True               2009   \n",
       "3               VWC                          NaN   True               1986   \n",
       "4               NaN                          NaN   True                  0   \n",
       "\n",
       "  extraction_type extraction_type_group extraction_type_class management  \\\n",
       "0         gravity               gravity               gravity        vwc   \n",
       "1         gravity               gravity               gravity        wug   \n",
       "2         gravity               gravity               gravity        vwc   \n",
       "3     submersible           submersible           submersible        vwc   \n",
       "4         gravity               gravity               gravity      other   \n",
       "\n",
       "  management_group         payment payment_type water_quality quality_group  \\\n",
       "0       user-group    pay annually     annually          soft          good   \n",
       "1       user-group       never pay    never pay          soft          good   \n",
       "2       user-group  pay per bucket   per bucket          soft          good   \n",
       "3       user-group       never pay    never pay          soft          good   \n",
       "4            other       never pay    never pay          soft          good   \n",
       "\n",
       "       quantity quantity_group                source           source_type  \\\n",
       "0        enough         enough                spring                spring   \n",
       "1  insufficient   insufficient  rainwater harvesting  rainwater harvesting   \n",
       "2        enough         enough                   dam                   dam   \n",
       "3           dry            dry           machine dbh              borehole   \n",
       "4      seasonal       seasonal  rainwater harvesting  rainwater harvesting   \n",
       "\n",
       "  source_class              waterpoint_type waterpoint_type_group  \\\n",
       "0  groundwater           communal standpipe    communal standpipe   \n",
       "1      surface           communal standpipe    communal standpipe   \n",
       "2      surface  communal standpipe multiple    communal standpipe   \n",
       "3  groundwater  communal standpipe multiple    communal standpipe   \n",
       "4      surface           communal standpipe    communal standpipe   \n",
       "\n",
       "     status_group  \n",
       "0      functional  \n",
       "1      functional  \n",
       "2      functional  \n",
       "3  non functional  \n",
       "4      functional  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Предобработка"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"panel panel-warning\">\n",
    "    <div class=\"panel-heading\">\n",
    "        <h3 class=\"panel-title\">Обратите внимание</h3> \n",
    "    </div>\n",
    "</div>\n",
    "\n",
    "Предобработка - опциональный блок, и у себя подготовить данные вы можете полностью по-своему.\n",
    "\n",
    "__Единственное замечание:__ если решите подготавливать данные самостоятельно, замените странную строку `\"other - mkulima/shinyanga\"` на просто `\"other\"`, так как в тесте только `\"other\"`.\n",
    "```python\n",
    "df.loc[df.extraction_type == 'other - mkulima/shinyanga', 'extraction_type'] = 'other'\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:45:57.750902",
     "start_time": "2017-07-25T11:45:57.745106"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def reduce_factor_levels(df, column_name, limit=None, top=None, name=None):\n",
    "    assert(limit is not None or top is not None), 'Specify limit ot top'\n",
    "    if top is None:\n",
    "        top = df[column_name].value_counts()[:limit].index\n",
    "    if name is None:\n",
    "        name = '%s_OTHER' % column_name\n",
    "    df.loc[~df[column_name].isin(top), column_name] = name\n",
    "    return top"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:45:58.259697",
     "start_time": "2017-07-25T11:45:58.140849"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "top = reduce_factor_levels(df, 'funder', 10)\n",
    "reduce_factor_levels(df_test, 'funder', top=top);\n",
    "\n",
    "top = reduce_factor_levels(df, 'installer', 10)\n",
    "reduce_factor_levels(df_test, 'installer', top=top);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:45:58.634685",
     "start_time": "2017-07-25T11:45:58.606891"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#drop = ['wpt_name', 'num_private', 'subvillage', 'region_code', 'district_code', 'lga', 'ward', 'recorded_by', 'scheme_name']\n",
    "\n",
    "drop = ['wpt_name', 'num_private', 'district_code', 'region_code', 'subvillage']  # \n",
    "\n",
    "df.drop(drop, axis=1, inplace=True)\n",
    "df_test.drop(drop, axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:45:59.021427",
     "start_time": "2017-07-25T11:45:58.999154"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df.loc[df.scheme_management == 'None', 'scheme_management'] = ''\n",
    "df.loc[df.scheme_management.isnull(), 'scheme_management'] = ''\n",
    "\n",
    "df_test.loc[df_test.scheme_management.isnull(), 'scheme_management'] = ''"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:45:59.399190",
     "start_time": "2017-07-25T11:45:59.391367"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df['construction_date_known'] = (df.construction_year > 0).astype(np.int32)\n",
    "df_test['construction_date_known'] = (df_test.construction_year > 0).astype(np.int32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:45:59.839112",
     "start_time": "2017-07-25T11:45:59.808318"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "min_year = df[df.construction_year > 0].construction_year.min() // 10 - 1\n",
    "\n",
    "df['construction_decade'] = df.construction_year // 10 - min_year\n",
    "df_test['construction_decade'] = df_test.construction_year // 10 - min_year\n",
    "\n",
    "df.loc[df.construction_decade < 0, 'construction_decade'] = 0\n",
    "df_test.loc[df_test.construction_decade < 0, 'construction_decade'] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:46:00.147440",
     "start_time": "2017-07-25T11:46:00.132709"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "top = reduce_factor_levels(df, 'construction_year', 20)\n",
    "reduce_factor_levels(df_test, 'construction_year', top=top);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:46:00.571709",
     "start_time": "2017-07-25T11:46:00.536218"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df.loc[df.extraction_type == 'other - mkulima/shinyanga', 'extraction_type'] = 'other'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:46:01.007007",
     "start_time": "2017-07-25T11:46:00.965509"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "heights = np.arange(-1, df.gps_height.max()+500, 500)\n",
    "height_labels = list(range(len(heights)-1))\n",
    "\n",
    "df['gps_height_rounded'] = pd.cut(df.gps_height, bins=heights, labels=height_labels)\n",
    "df_test['gps_height_rounded'] = pd.cut(df_test.gps_height, bins=heights, labels=height_labels)\n",
    "\n",
    "#df.drop(['gps_height'], axis=1, inplace=True)\n",
    "#df_test.drop(['gps_height'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:46:01.405938",
     "start_time": "2017-07-25T11:46:01.371469"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#pops = np.arange(-1, df.population.max()+500, 500)\n",
    "#pops_labels = list(range(len(pops)-1))\n",
    "\n",
    "#df['pop_rounded'] = pd.cut(df.population, bins=pops, labels=pops_labels)\n",
    "#df_test['pop_rounded'] = pd.cut(df_test.population, bins=pops, labels=pops_labels)\n",
    "\n",
    "#df.drop(['population'], axis=1, inplace=True)\n",
    "#df_test.drop(['population'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:46:02.914167",
     "start_time": "2017-07-25T11:46:02.891364"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#df.drop(['date_recorded'], axis=1, inplace=True)\n",
    "#df_test.drop(['date_recorded'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:46:03.324955",
     "start_time": "2017-07-25T11:46:03.303671"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df.public_meeting.fillna(True, inplace=True)\n",
    "df_test.public_meeting.fillna(True, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:46:03.746867",
     "start_time": "2017-07-25T11:46:03.725670"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df.permit.fillna(True, inplace=True)\n",
    "df_test.permit.fillna(True, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:46:04.760107",
     "start_time": "2017-07-25T11:46:04.754408"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df.gps_height_rounded.fillna(0, inplace=True)\n",
    "df_test.gps_height_rounded.fillna(0, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Визуализация"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "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>id</th>\n",
       "      <th>amount_tsh</th>\n",
       "      <th>date_recorded</th>\n",
       "      <th>funder</th>\n",
       "      <th>installer</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>basin</th>\n",
       "      <th>region</th>\n",
       "      <th>public_meeting</th>\n",
       "      <th>scheme_management</th>\n",
       "      <th>permit</th>\n",
       "      <th>construction_year</th>\n",
       "      <th>extraction_type</th>\n",
       "      <th>extraction_type_group</th>\n",
       "      <th>extraction_type_class</th>\n",
       "      <th>management</th>\n",
       "      <th>management_group</th>\n",
       "      <th>payment</th>\n",
       "      <th>payment_type</th>\n",
       "      <th>water_quality</th>\n",
       "      <th>quality_group</th>\n",
       "      <th>quantity</th>\n",
       "      <th>quantity_group</th>\n",
       "      <th>source</th>\n",
       "      <th>source_type</th>\n",
       "      <th>source_class</th>\n",
       "      <th>waterpoint_type</th>\n",
       "      <th>waterpoint_type_group</th>\n",
       "      <th>status_group</th>\n",
       "      <th>construction_date_known</th>\n",
       "      <th>construction_decade</th>\n",
       "      <th>gps_height_rounded</th>\n",
       "      <th>pop_rounded</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>69572</td>\n",
       "      <td>6000.0</td>\n",
       "      <td>2011-03-14</td>\n",
       "      <td>funder_OTHER</td>\n",
       "      <td>installer_OTHER</td>\n",
       "      <td>34.938093</td>\n",
       "      <td>-9.856322</td>\n",
       "      <td>Lake Nyasa</td>\n",
       "      <td>Iringa</td>\n",
       "      <td>True</td>\n",
       "      <td>VWC</td>\n",
       "      <td>False</td>\n",
       "      <td>1999</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>vwc</td>\n",
       "      <td>user-group</td>\n",
       "      <td>pay annually</td>\n",
       "      <td>annually</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>enough</td>\n",
       "      <td>spring</td>\n",
       "      <td>spring</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>functional</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8776</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2013-03-06</td>\n",
       "      <td>funder_OTHER</td>\n",
       "      <td>installer_OTHER</td>\n",
       "      <td>34.698766</td>\n",
       "      <td>-2.147466</td>\n",
       "      <td>Lake Victoria</td>\n",
       "      <td>Mara</td>\n",
       "      <td>True</td>\n",
       "      <td>Other</td>\n",
       "      <td>True</td>\n",
       "      <td>2010</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>wug</td>\n",
       "      <td>user-group</td>\n",
       "      <td>never pay</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>insufficient</td>\n",
       "      <td>insufficient</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>functional</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>34310</td>\n",
       "      <td>25.0</td>\n",
       "      <td>2013-02-25</td>\n",
       "      <td>funder_OTHER</td>\n",
       "      <td>installer_OTHER</td>\n",
       "      <td>37.460664</td>\n",
       "      <td>-3.821329</td>\n",
       "      <td>Pangani</td>\n",
       "      <td>Manyara</td>\n",
       "      <td>True</td>\n",
       "      <td>VWC</td>\n",
       "      <td>True</td>\n",
       "      <td>2009</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>vwc</td>\n",
       "      <td>user-group</td>\n",
       "      <td>pay per bucket</td>\n",
       "      <td>per bucket</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>enough</td>\n",
       "      <td>dam</td>\n",
       "      <td>dam</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe multiple</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>functional</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>67743</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2013-01-28</td>\n",
       "      <td>Unicef</td>\n",
       "      <td>installer_OTHER</td>\n",
       "      <td>38.486161</td>\n",
       "      <td>-11.155298</td>\n",
       "      <td>Ruvuma / Southern Coast</td>\n",
       "      <td>Mtwara</td>\n",
       "      <td>True</td>\n",
       "      <td>VWC</td>\n",
       "      <td>True</td>\n",
       "      <td>construction_year_OTHER</td>\n",
       "      <td>submersible</td>\n",
       "      <td>submersible</td>\n",
       "      <td>submersible</td>\n",
       "      <td>vwc</td>\n",
       "      <td>user-group</td>\n",
       "      <td>never pay</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>dry</td>\n",
       "      <td>dry</td>\n",
       "      <td>machine dbh</td>\n",
       "      <td>borehole</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe multiple</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>non functional</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>19728</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2011-07-13</td>\n",
       "      <td>funder_OTHER</td>\n",
       "      <td>installer_OTHER</td>\n",
       "      <td>31.130847</td>\n",
       "      <td>-1.825359</td>\n",
       "      <td>Lake Victoria</td>\n",
       "      <td>Kagera</td>\n",
       "      <td>True</td>\n",
       "      <td></td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>other</td>\n",
       "      <td>other</td>\n",
       "      <td>never pay</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>seasonal</td>\n",
       "      <td>seasonal</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>functional</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id  amount_tsh date_recorded        funder        installer  longitude  \\\n",
       "0  69572      6000.0    2011-03-14  funder_OTHER  installer_OTHER  34.938093   \n",
       "1   8776         0.0    2013-03-06  funder_OTHER  installer_OTHER  34.698766   \n",
       "2  34310        25.0    2013-02-25  funder_OTHER  installer_OTHER  37.460664   \n",
       "3  67743         0.0    2013-01-28        Unicef  installer_OTHER  38.486161   \n",
       "4  19728         0.0    2011-07-13  funder_OTHER  installer_OTHER  31.130847   \n",
       "\n",
       "    latitude                    basin   region  public_meeting  \\\n",
       "0  -9.856322               Lake Nyasa   Iringa            True   \n",
       "1  -2.147466            Lake Victoria     Mara            True   \n",
       "2  -3.821329                  Pangani  Manyara            True   \n",
       "3 -11.155298  Ruvuma / Southern Coast   Mtwara            True   \n",
       "4  -1.825359            Lake Victoria   Kagera            True   \n",
       "\n",
       "  scheme_management  permit        construction_year extraction_type  \\\n",
       "0               VWC   False                     1999         gravity   \n",
       "1             Other    True                     2010         gravity   \n",
       "2               VWC    True                     2009         gravity   \n",
       "3               VWC    True  construction_year_OTHER     submersible   \n",
       "4                      True                        0         gravity   \n",
       "\n",
       "  extraction_type_group extraction_type_class management management_group  \\\n",
       "0               gravity               gravity        vwc       user-group   \n",
       "1               gravity               gravity        wug       user-group   \n",
       "2               gravity               gravity        vwc       user-group   \n",
       "3           submersible           submersible        vwc       user-group   \n",
       "4               gravity               gravity      other            other   \n",
       "\n",
       "          payment payment_type water_quality quality_group      quantity  \\\n",
       "0    pay annually     annually          soft          good        enough   \n",
       "1       never pay    never pay          soft          good  insufficient   \n",
       "2  pay per bucket   per bucket          soft          good        enough   \n",
       "3       never pay    never pay          soft          good           dry   \n",
       "4       never pay    never pay          soft          good      seasonal   \n",
       "\n",
       "  quantity_group                source           source_type source_class  \\\n",
       "0         enough                spring                spring  groundwater   \n",
       "1   insufficient  rainwater harvesting  rainwater harvesting      surface   \n",
       "2         enough                   dam                   dam      surface   \n",
       "3            dry           machine dbh              borehole  groundwater   \n",
       "4       seasonal  rainwater harvesting  rainwater harvesting      surface   \n",
       "\n",
       "               waterpoint_type waterpoint_type_group    status_group  \\\n",
       "0           communal standpipe    communal standpipe      functional   \n",
       "1           communal standpipe    communal standpipe      functional   \n",
       "2  communal standpipe multiple    communal standpipe      functional   \n",
       "3  communal standpipe multiple    communal standpipe  non functional   \n",
       "4           communal standpipe    communal standpipe      functional   \n",
       "\n",
       "   construction_date_known  construction_decade gps_height_rounded pop_rounded  \n",
       "0                        1                    4                  2           0  \n",
       "1                        1                    6                  2           0  \n",
       "2                        1                    5                  1           0  \n",
       "3                        1                    3                  0           0  \n",
       "4                        0                    0                  0           0  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Визуализируем качество воды в разных регионах"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "good        50818\n",
       "salty        5195\n",
       "unknown      1876\n",
       "milky         804\n",
       "colored       490\n",
       "fluoride      217\n",
       "Name: quality_group, dtype: int64"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.quality_group.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABDEAAAGrCAYAAADZzwwPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x29cacbfa358>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "quality_groupInRegions = df.groupby('region')['quality_group'].value_counts().to_dict()\n",
    "\n",
    "results = pd.DataFrame(data = quality_groupInRegions, index=[0]).stack().fillna(0).transpose()\n",
    "results.columns = pd.Index(['good', 'salty', 'unknown', 'milky', 'colored', 'fluoride'])\n",
    "results['total'] = results.good + results.salty + results.unknown + results.milky + results.colored + results.fluoride\n",
    "results.sort_values(by='good', ascending=False, inplace=True)\n",
    "results[['good', 'salty', 'unknown', 'milky', 'colored', 'fluoride']].plot(kind='bar', stacked=True, rot=45);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Визуализируем геоданные о текущем статусе водяных насосов"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:46:26.438343",
     "start_time": "2017-07-25T11:46:26.435545"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from folium import Map, CircleMarker\n",
    "import colorsys\n",
    "\n",
    "#Просто карта\n",
    "tanzania_map = Map(location=(-2.147466, 34.698766), tiles='Mapbox Bright', zoom_start=6)\n",
    "#tanzania_map"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "functional                 32259\n",
       "non functional             22824\n",
       "functional needs repair     4317\n",
       "Name: status_group, dtype: int64"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#в качестве target сделаем текущий статус\n",
    "df.status_group.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "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>id</th>\n",
       "      <th>amount_tsh</th>\n",
       "      <th>funder</th>\n",
       "      <th>installer</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>basin</th>\n",
       "      <th>region</th>\n",
       "      <th>public_meeting</th>\n",
       "      <th>scheme_management</th>\n",
       "      <th>permit</th>\n",
       "      <th>construction_year</th>\n",
       "      <th>extraction_type</th>\n",
       "      <th>extraction_type_group</th>\n",
       "      <th>extraction_type_class</th>\n",
       "      <th>management</th>\n",
       "      <th>management_group</th>\n",
       "      <th>payment</th>\n",
       "      <th>payment_type</th>\n",
       "      <th>water_quality</th>\n",
       "      <th>quality_group</th>\n",
       "      <th>quantity</th>\n",
       "      <th>quantity_group</th>\n",
       "      <th>source</th>\n",
       "      <th>source_type</th>\n",
       "      <th>source_class</th>\n",
       "      <th>waterpoint_type</th>\n",
       "      <th>waterpoint_type_group</th>\n",
       "      <th>status_group</th>\n",
       "      <th>construction_date_known</th>\n",
       "      <th>construction_decade</th>\n",
       "      <th>gps_height_rounded</th>\n",
       "      <th>pop_rounded</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>69572</td>\n",
       "      <td>6000.0</td>\n",
       "      <td>funder_OTHER</td>\n",
       "      <td>installer_OTHER</td>\n",
       "      <td>34.938093</td>\n",
       "      <td>-9.856322</td>\n",
       "      <td>Lake Nyasa</td>\n",
       "      <td>Iringa</td>\n",
       "      <td>True</td>\n",
       "      <td>VWC</td>\n",
       "      <td>False</td>\n",
       "      <td>1999</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>vwc</td>\n",
       "      <td>user-group</td>\n",
       "      <td>pay annually</td>\n",
       "      <td>annually</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>enough</td>\n",
       "      <td>spring</td>\n",
       "      <td>spring</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>functional</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8776</td>\n",
       "      <td>0.0</td>\n",
       "      <td>funder_OTHER</td>\n",
       "      <td>installer_OTHER</td>\n",
       "      <td>34.698766</td>\n",
       "      <td>-2.147466</td>\n",
       "      <td>Lake Victoria</td>\n",
       "      <td>Mara</td>\n",
       "      <td>True</td>\n",
       "      <td>Other</td>\n",
       "      <td>True</td>\n",
       "      <td>2010</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>wug</td>\n",
       "      <td>user-group</td>\n",
       "      <td>never pay</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>insufficient</td>\n",
       "      <td>insufficient</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>functional</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>34310</td>\n",
       "      <td>25.0</td>\n",
       "      <td>funder_OTHER</td>\n",
       "      <td>installer_OTHER</td>\n",
       "      <td>37.460664</td>\n",
       "      <td>-3.821329</td>\n",
       "      <td>Pangani</td>\n",
       "      <td>Manyara</td>\n",
       "      <td>True</td>\n",
       "      <td>VWC</td>\n",
       "      <td>True</td>\n",
       "      <td>2009</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>vwc</td>\n",
       "      <td>user-group</td>\n",
       "      <td>pay per bucket</td>\n",
       "      <td>per bucket</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>enough</td>\n",
       "      <td>dam</td>\n",
       "      <td>dam</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe multiple</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>functional</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>67743</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Unicef</td>\n",
       "      <td>installer_OTHER</td>\n",
       "      <td>38.486161</td>\n",
       "      <td>-11.155298</td>\n",
       "      <td>Ruvuma / Southern Coast</td>\n",
       "      <td>Mtwara</td>\n",
       "      <td>True</td>\n",
       "      <td>VWC</td>\n",
       "      <td>True</td>\n",
       "      <td>construction_year_OTHER</td>\n",
       "      <td>submersible</td>\n",
       "      <td>submersible</td>\n",
       "      <td>submersible</td>\n",
       "      <td>vwc</td>\n",
       "      <td>user-group</td>\n",
       "      <td>never pay</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>dry</td>\n",
       "      <td>dry</td>\n",
       "      <td>machine dbh</td>\n",
       "      <td>borehole</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe multiple</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>non functional</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>19728</td>\n",
       "      <td>0.0</td>\n",
       "      <td>funder_OTHER</td>\n",
       "      <td>installer_OTHER</td>\n",
       "      <td>31.130847</td>\n",
       "      <td>-1.825359</td>\n",
       "      <td>Lake Victoria</td>\n",
       "      <td>Kagera</td>\n",
       "      <td>True</td>\n",
       "      <td></td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>other</td>\n",
       "      <td>other</td>\n",
       "      <td>never pay</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>seasonal</td>\n",
       "      <td>seasonal</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>functional</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id  amount_tsh        funder        installer  longitude   latitude  \\\n",
       "0  69572      6000.0  funder_OTHER  installer_OTHER  34.938093  -9.856322   \n",
       "1   8776         0.0  funder_OTHER  installer_OTHER  34.698766  -2.147466   \n",
       "2  34310        25.0  funder_OTHER  installer_OTHER  37.460664  -3.821329   \n",
       "3  67743         0.0        Unicef  installer_OTHER  38.486161 -11.155298   \n",
       "4  19728         0.0  funder_OTHER  installer_OTHER  31.130847  -1.825359   \n",
       "\n",
       "                     basin   region  public_meeting scheme_management  permit  \\\n",
       "0               Lake Nyasa   Iringa            True               VWC   False   \n",
       "1            Lake Victoria     Mara            True             Other    True   \n",
       "2                  Pangani  Manyara            True               VWC    True   \n",
       "3  Ruvuma / Southern Coast   Mtwara            True               VWC    True   \n",
       "4            Lake Victoria   Kagera            True                      True   \n",
       "\n",
       "         construction_year extraction_type extraction_type_group  \\\n",
       "0                     1999         gravity               gravity   \n",
       "1                     2010         gravity               gravity   \n",
       "2                     2009         gravity               gravity   \n",
       "3  construction_year_OTHER     submersible           submersible   \n",
       "4                        0         gravity               gravity   \n",
       "\n",
       "  extraction_type_class management management_group         payment  \\\n",
       "0               gravity        vwc       user-group    pay annually   \n",
       "1               gravity        wug       user-group       never pay   \n",
       "2               gravity        vwc       user-group  pay per bucket   \n",
       "3           submersible        vwc       user-group       never pay   \n",
       "4               gravity      other            other       never pay   \n",
       "\n",
       "  payment_type water_quality quality_group      quantity quantity_group  \\\n",
       "0     annually          soft          good        enough         enough   \n",
       "1    never pay          soft          good  insufficient   insufficient   \n",
       "2   per bucket          soft          good        enough         enough   \n",
       "3    never pay          soft          good           dry            dry   \n",
       "4    never pay          soft          good      seasonal       seasonal   \n",
       "\n",
       "                 source           source_type source_class  \\\n",
       "0                spring                spring  groundwater   \n",
       "1  rainwater harvesting  rainwater harvesting      surface   \n",
       "2                   dam                   dam      surface   \n",
       "3           machine dbh              borehole  groundwater   \n",
       "4  rainwater harvesting  rainwater harvesting      surface   \n",
       "\n",
       "               waterpoint_type waterpoint_type_group    status_group  \\\n",
       "0           communal standpipe    communal standpipe      functional   \n",
       "1           communal standpipe    communal standpipe      functional   \n",
       "2  communal standpipe multiple    communal standpipe      functional   \n",
       "3  communal standpipe multiple    communal standpipe  non functional   \n",
       "4           communal standpipe    communal standpipe      functional   \n",
       "\n",
       "   construction_date_known  construction_decade gps_height_rounded  \\\n",
       "0                        1                    4                  2   \n",
       "1                        1                    6                  2   \n",
       "2                        1                    5                  1   \n",
       "3                        1                    3                  0   \n",
       "4                        0                    0                  0   \n",
       "\n",
       "  pop_rounded  target  \n",
       "0           0       2  \n",
       "1           0       2  \n",
       "2           0       2  \n",
       "3           0       0  \n",
       "4           0       2  "
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['target'] = 0 #non functional\n",
    "df.loc[df.status_group == 'functional needs repair', 'target'] = 1\n",
    "df.loc[df.status_group == 'functional', 'target'] = 2\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Карта с отмеченными источниками\n",
    "get_radius = lambda x: (x - min_)/(max_ - min_)*7 + 3\n",
    "\n",
    "rgbhex = lambda rgb: '#'+\"\".join(\"%02X\" % i for i in rgb)\n",
    "get_fill_color = lambda x: rgbhex(tuple(int(i * 255) for i in \\\n",
    "                                   colorsys.hsv_to_rgb(x/max_*120.0/360.0, 0.56, 0.84)))\n",
    "get_border_color = lambda x: rgbhex(tuple(int(i * 255) for i in \\\n",
    "                                   colorsys.hsv_to_rgb(x/max_*120.0/360.0, 0.78, 0.36)))\n",
    "\n",
    "add_marker = lambda lat, lon, target: \\\n",
    "    CircleMarker((lat, lon), \n",
    "                 radius     = get_radius(target), \n",
    "                 color      = get_border_color(target),\n",
    "                 fill_color = get_fill_color(target), \n",
    "                 popup='Lat: %.3f; Lon: %.3f' % (lat, lon),\n",
    "                    )\\\n",
    "    .add_to(tanzania_map)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "min_, max_ = df[['target']].describe().loc['min'][0], df[['target']].describe().loc['max'][0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df.sample(n=1000).apply(lambda row: add_marker(row['latitude'], row['longitude'], row['target']), axis=1);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true,
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "#tanzania_map"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "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>id</th>\n",
       "      <th>amount_tsh</th>\n",
       "      <th>funder</th>\n",
       "      <th>installer</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>basin</th>\n",
       "      <th>region</th>\n",
       "      <th>public_meeting</th>\n",
       "      <th>scheme_management</th>\n",
       "      <th>permit</th>\n",
       "      <th>construction_year</th>\n",
       "      <th>extraction_type</th>\n",
       "      <th>extraction_type_group</th>\n",
       "      <th>extraction_type_class</th>\n",
       "      <th>management</th>\n",
       "      <th>management_group</th>\n",
       "      <th>payment</th>\n",
       "      <th>payment_type</th>\n",
       "      <th>water_quality</th>\n",
       "      <th>quality_group</th>\n",
       "      <th>quantity</th>\n",
       "      <th>quantity_group</th>\n",
       "      <th>source</th>\n",
       "      <th>source_type</th>\n",
       "      <th>source_class</th>\n",
       "      <th>waterpoint_type</th>\n",
       "      <th>waterpoint_type_group</th>\n",
       "      <th>status_group</th>\n",
       "      <th>construction_date_known</th>\n",
       "      <th>construction_decade</th>\n",
       "      <th>gps_height_rounded</th>\n",
       "      <th>pop_rounded</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>69572</td>\n",
       "      <td>6000.0</td>\n",
       "      <td>funder_OTHER</td>\n",
       "      <td>installer_OTHER</td>\n",
       "      <td>34.938093</td>\n",
       "      <td>-9.856322</td>\n",
       "      <td>Lake Nyasa</td>\n",
       "      <td>Iringa</td>\n",
       "      <td>True</td>\n",
       "      <td>VWC</td>\n",
       "      <td>False</td>\n",
       "      <td>1999</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>vwc</td>\n",
       "      <td>user-group</td>\n",
       "      <td>pay annually</td>\n",
       "      <td>annually</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>enough</td>\n",
       "      <td>spring</td>\n",
       "      <td>spring</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>functional</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8776</td>\n",
       "      <td>0.0</td>\n",
       "      <td>funder_OTHER</td>\n",
       "      <td>installer_OTHER</td>\n",
       "      <td>34.698766</td>\n",
       "      <td>-2.147466</td>\n",
       "      <td>Lake Victoria</td>\n",
       "      <td>Mara</td>\n",
       "      <td>True</td>\n",
       "      <td>Other</td>\n",
       "      <td>True</td>\n",
       "      <td>2010</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>wug</td>\n",
       "      <td>user-group</td>\n",
       "      <td>never pay</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>insufficient</td>\n",
       "      <td>insufficient</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>functional</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>34310</td>\n",
       "      <td>25.0</td>\n",
       "      <td>funder_OTHER</td>\n",
       "      <td>installer_OTHER</td>\n",
       "      <td>37.460664</td>\n",
       "      <td>-3.821329</td>\n",
       "      <td>Pangani</td>\n",
       "      <td>Manyara</td>\n",
       "      <td>True</td>\n",
       "      <td>VWC</td>\n",
       "      <td>True</td>\n",
       "      <td>2009</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>vwc</td>\n",
       "      <td>user-group</td>\n",
       "      <td>pay per bucket</td>\n",
       "      <td>per bucket</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>enough</td>\n",
       "      <td>dam</td>\n",
       "      <td>dam</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe multiple</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>functional</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>67743</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Unicef</td>\n",
       "      <td>installer_OTHER</td>\n",
       "      <td>38.486161</td>\n",
       "      <td>-11.155298</td>\n",
       "      <td>Ruvuma / Southern Coast</td>\n",
       "      <td>Mtwara</td>\n",
       "      <td>True</td>\n",
       "      <td>VWC</td>\n",
       "      <td>True</td>\n",
       "      <td>construction_year_OTHER</td>\n",
       "      <td>submersible</td>\n",
       "      <td>submersible</td>\n",
       "      <td>submersible</td>\n",
       "      <td>vwc</td>\n",
       "      <td>user-group</td>\n",
       "      <td>never pay</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>dry</td>\n",
       "      <td>dry</td>\n",
       "      <td>machine dbh</td>\n",
       "      <td>borehole</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe multiple</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>non functional</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>19728</td>\n",
       "      <td>0.0</td>\n",
       "      <td>funder_OTHER</td>\n",
       "      <td>installer_OTHER</td>\n",
       "      <td>31.130847</td>\n",
       "      <td>-1.825359</td>\n",
       "      <td>Lake Victoria</td>\n",
       "      <td>Kagera</td>\n",
       "      <td>True</td>\n",
       "      <td></td>\n",
       "      <td>True</td>\n",
       "      <td>0</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>other</td>\n",
       "      <td>other</td>\n",
       "      <td>never pay</td>\n",
       "      <td>never pay</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>seasonal</td>\n",
       "      <td>seasonal</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>rainwater harvesting</td>\n",
       "      <td>surface</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>functional</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id  amount_tsh        funder        installer  longitude   latitude  \\\n",
       "0  69572      6000.0  funder_OTHER  installer_OTHER  34.938093  -9.856322   \n",
       "1   8776         0.0  funder_OTHER  installer_OTHER  34.698766  -2.147466   \n",
       "2  34310        25.0  funder_OTHER  installer_OTHER  37.460664  -3.821329   \n",
       "3  67743         0.0        Unicef  installer_OTHER  38.486161 -11.155298   \n",
       "4  19728         0.0  funder_OTHER  installer_OTHER  31.130847  -1.825359   \n",
       "\n",
       "                     basin   region  public_meeting scheme_management  permit  \\\n",
       "0               Lake Nyasa   Iringa            True               VWC   False   \n",
       "1            Lake Victoria     Mara            True             Other    True   \n",
       "2                  Pangani  Manyara            True               VWC    True   \n",
       "3  Ruvuma / Southern Coast   Mtwara            True               VWC    True   \n",
       "4            Lake Victoria   Kagera            True                      True   \n",
       "\n",
       "         construction_year extraction_type extraction_type_group  \\\n",
       "0                     1999         gravity               gravity   \n",
       "1                     2010         gravity               gravity   \n",
       "2                     2009         gravity               gravity   \n",
       "3  construction_year_OTHER     submersible           submersible   \n",
       "4                        0         gravity               gravity   \n",
       "\n",
       "  extraction_type_class management management_group         payment  \\\n",
       "0               gravity        vwc       user-group    pay annually   \n",
       "1               gravity        wug       user-group       never pay   \n",
       "2               gravity        vwc       user-group  pay per bucket   \n",
       "3           submersible        vwc       user-group       never pay   \n",
       "4               gravity      other            other       never pay   \n",
       "\n",
       "  payment_type water_quality quality_group      quantity quantity_group  \\\n",
       "0     annually          soft          good        enough         enough   \n",
       "1    never pay          soft          good  insufficient   insufficient   \n",
       "2   per bucket          soft          good        enough         enough   \n",
       "3    never pay          soft          good           dry            dry   \n",
       "4    never pay          soft          good      seasonal       seasonal   \n",
       "\n",
       "                 source           source_type source_class  \\\n",
       "0                spring                spring  groundwater   \n",
       "1  rainwater harvesting  rainwater harvesting      surface   \n",
       "2                   dam                   dam      surface   \n",
       "3           machine dbh              borehole  groundwater   \n",
       "4  rainwater harvesting  rainwater harvesting      surface   \n",
       "\n",
       "               waterpoint_type waterpoint_type_group    status_group  \\\n",
       "0           communal standpipe    communal standpipe      functional   \n",
       "1           communal standpipe    communal standpipe      functional   \n",
       "2  communal standpipe multiple    communal standpipe      functional   \n",
       "3  communal standpipe multiple    communal standpipe  non functional   \n",
       "4           communal standpipe    communal standpipe      functional   \n",
       "\n",
       "   construction_date_known  construction_decade gps_height_rounded pop_rounded  \n",
       "0                        1                    4                  2           0  \n",
       "1                        1                    6                  2           0  \n",
       "2                        1                    5                  1           0  \n",
       "3                        1                    3                  0           0  \n",
       "4                        0                    0                  0           0  "
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df.drop(['target'], axis=1)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Модели"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:46:37.396648",
     "start_time": "2017-07-25T11:46:37.378335"
    },
    "collapsed": true,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "X, y, X_test = df.drop(['id', 'status_group'], axis=1), \\\n",
    "               df.status_group, \\\n",
    "               df_test.drop(['id'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:46:38.445433",
     "start_time": "2017-07-25T11:46:38.407894"
    }
   },
   "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>amount_tsh</th>\n",
       "      <th>date_recorded</th>\n",
       "      <th>funder</th>\n",
       "      <th>gps_height</th>\n",
       "      <th>installer</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>basin</th>\n",
       "      <th>region</th>\n",
       "      <th>lga</th>\n",
       "      <th>ward</th>\n",
       "      <th>population</th>\n",
       "      <th>public_meeting</th>\n",
       "      <th>recorded_by</th>\n",
       "      <th>scheme_management</th>\n",
       "      <th>scheme_name</th>\n",
       "      <th>permit</th>\n",
       "      <th>construction_year</th>\n",
       "      <th>extraction_type</th>\n",
       "      <th>extraction_type_group</th>\n",
       "      <th>extraction_type_class</th>\n",
       "      <th>management</th>\n",
       "      <th>management_group</th>\n",
       "      <th>payment</th>\n",
       "      <th>payment_type</th>\n",
       "      <th>water_quality</th>\n",
       "      <th>quality_group</th>\n",
       "      <th>quantity</th>\n",
       "      <th>quantity_group</th>\n",
       "      <th>source</th>\n",
       "      <th>source_type</th>\n",
       "      <th>source_class</th>\n",
       "      <th>waterpoint_type</th>\n",
       "      <th>waterpoint_type_group</th>\n",
       "      <th>construction_date_known</th>\n",
       "      <th>construction_decade</th>\n",
       "      <th>gps_height_rounded</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6000.0</td>\n",
       "      <td>2011-03-14</td>\n",
       "      <td>funder_OTHER</td>\n",
       "      <td>1390</td>\n",
       "      <td>installer_OTHER</td>\n",
       "      <td>34.938093</td>\n",
       "      <td>-9.856322</td>\n",
       "      <td>Lake Nyasa</td>\n",
       "      <td>Iringa</td>\n",
       "      <td>Ludewa</td>\n",
       "      <td>Mundindi</td>\n",
       "      <td>109</td>\n",
       "      <td>True</td>\n",
       "      <td>GeoData Consultants Ltd</td>\n",
       "      <td>VWC</td>\n",
       "      <td>Roman</td>\n",
       "      <td>False</td>\n",
       "      <td>1999</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>gravity</td>\n",
       "      <td>vwc</td>\n",
       "      <td>user-group</td>\n",
       "      <td>pay annually</td>\n",
       "      <td>annually</td>\n",
       "      <td>soft</td>\n",
       "      <td>good</td>\n",
       "      <td>enough</td>\n",
       "      <td>enough</td>\n",
       "      <td>spring</td>\n",
       "      <td>spring</td>\n",
       "      <td>groundwater</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>communal standpipe</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   amount_tsh date_recorded        funder  gps_height        installer  \\\n",
       "0      6000.0    2011-03-14  funder_OTHER        1390  installer_OTHER   \n",
       "\n",
       "   longitude  latitude       basin  region     lga      ward  population  \\\n",
       "0  34.938093 -9.856322  Lake Nyasa  Iringa  Ludewa  Mundindi         109   \n",
       "\n",
       "   public_meeting              recorded_by scheme_management scheme_name  \\\n",
       "0            True  GeoData Consultants Ltd               VWC       Roman   \n",
       "\n",
       "   permit construction_year extraction_type extraction_type_group  \\\n",
       "0   False              1999         gravity               gravity   \n",
       "\n",
       "  extraction_type_class management management_group       payment  \\\n",
       "0               gravity        vwc       user-group  pay annually   \n",
       "\n",
       "  payment_type water_quality quality_group quantity quantity_group  source  \\\n",
       "0     annually          soft          good   enough         enough  spring   \n",
       "\n",
       "  source_type source_class     waterpoint_type waterpoint_type_group  \\\n",
       "0      spring  groundwater  communal standpipe    communal standpipe   \n",
       "\n",
       "   construction_date_known  construction_decade gps_height_rounded  \n",
       "0                        1                    4                  2  "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.head(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"panel panel-warning\">\n",
    "    <div class=\"panel-heading\">\n",
    "        <h3 class=\"panel-title\">Обратите внимание</h3> \n",
    "    </div>\n",
    "</div>\n",
    "\n",
    "Вот эта функция ниже - опять мои штуки-дрюки, и можно кодировать данные по-своему."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:46:58.463355",
     "start_time": "2017-07-25T11:46:58.450572"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def prepare(X_train, X_test):\n",
    "    from sklearn.preprocessing import StandardScaler\n",
    "    from sklearn.feature_extraction import DictVectorizer\n",
    "    \n",
    "    objects = X_train.select_dtypes(include=['O']).columns.values\n",
    "    numeric = X_train.select_dtypes(exclude=['O']).columns.values\n",
    "    \n",
    "    dv = DictVectorizer(sparse=False)\n",
    "    data_encoded_tr = dv.fit_transform(X_train[objects].to_dict(orient='records'))\n",
    "    data_encoded_ts = dv.transform(X_test[objects].to_dict(orient='records'))\n",
    "\n",
    "    ss = StandardScaler()\n",
    "    data_scaled_tr = ss.fit_transform(X_train[numeric])\n",
    "    data_scaled_ts = ss.transform(X_test[numeric])\n",
    "    \n",
    "    train = np.hstack((data_encoded_tr, data_scaled_tr))\n",
    "    test  = np.hstack((data_encoded_ts, data_scaled_ts))\n",
    "    return train, test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:47:05.391332",
     "start_time": "2017-07-25T11:47:00.997635"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "x_train, x_test = prepare(X, X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:47:05.400313",
     "start_time": "2017-07-25T11:47:05.393333"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import LabelEncoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-25T11:47:05.450353",
     "start_time": "2017-07-25T11:47:05.402339"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "y_encoder = LabelEncoder()\n",
    "y = y_encoder.fit_transform(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(59400, 5472)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(14850, 5472)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"panel panel-info\" style=\"margin: 50px 0 0 0\">\n",
    "    <div class=\"panel-heading\">\n",
    "        <h3 class=\"panel-title\">Задание 1.</h3> \n",
    "    </div>\n",
    "</div>\n",
    "\n",
    "Возьмите тетрадку с сегодняшнего занятия и, руководствуясь советами по настройке, заделайте лучший GBM в мире! Не забудьте отправлять результаты на drivendata и хвастаться в чате о результатах."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "SEED = 1234\n",
    "np.random.seed = SEED"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "from sklearn.model_selection import cross_val_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 258,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.69139730639730634, 0.0025182844378876025)"
      ]
     },
     "execution_count": 258,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf = GradientBoostingClassifier(random_state=SEED, n_estimators=10, learning_rate=0.01,subsample=0.8, max_depth=4) \n",
    "scores = cross_val_score(clf, x_train, y)\n",
    "np.mean(scores), 2*np.std(scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 259,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean score: 0.691397306397\n"
     ]
    }
   ],
   "source": [
    "clf = clf.fit(x_train, y)\n",
    "print('Mean score:', scores.mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 260,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, ..., 0, 0, 2], dtype=int64)"
      ]
     },
     "execution_count": 260,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_te = clf.predict(x_test)\n",
    "y_te"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 261,
   "metadata": {},
   "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>id</th>\n",
       "      <th>status_group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>50785</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>51630</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>17168</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>45559</td>\n",
       "      <td>non functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>49871</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id    status_group\n",
       "0  50785      functional\n",
       "1  51630      functional\n",
       "2  17168      functional\n",
       "3  45559  non functional\n",
       "4  49871      functional"
      ]
     },
     "execution_count": 261,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ans_nn2 = pd.DataFrame({'id': df_test['id'], 'status_group': y_encoder.inverse_transform(y_te)})\n",
    "ans_nn2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 262,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "ans_nn.to_csv('ans_gbm.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"panel panel-info\" style=\"margin: 50px 0 0 0\">\n",
    "    <div class=\"panel-heading\">\n",
    "        <h3 class=\"panel-title\">Задание 2.</h3> \n",
    "    </div>\n",
    "</div>\n",
    "\n",
    "Выберите любой из сторонних фреймворков по своему усмотрению:\n",
    "* XGBoost\n",
    "* LightGBM\n",
    "* H2O\n",
    "* CatBoost\n",
    "\n",
    "Установите, прокачайте его, побейте GBM от sklearn."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Опробуем lightgbm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-24T21:56:26.511616",
     "start_time": "2017-07-24T21:56:26.509076"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from lightgbm import LGBMClassifier\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.ensemble import VotingClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.80742424242424249, 0.0041631196002892687)"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#clf2 = LGBMClassifier(max_bin=475,learning_rate=0.13,n_estimators=140,num_leaves=131)\n",
    "#clf2 = LGBMClassifier(max_bin=400,learning_rate=0.13,n_estimators=140,num_leaves=131)\n",
    "\n",
    "clf2 = LGBMClassifier(max_bin=400,learning_rate=0.134,n_estimators=151,num_leaves=131)\n",
    " \n",
    "scores = cross_val_score(clf2, x_train, y)\n",
    "np.mean(scores), 2*np.std(scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "clf2 = clf2.fit(x_train, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, ..., 0, 0, 2], dtype=int64)"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_te = clf2.predict(x_test)\n",
    "y_te"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "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>id</th>\n",
       "      <th>status_group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>50785</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>51630</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>17168</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>45559</td>\n",
       "      <td>non functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>49871</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id    status_group\n",
       "0  50785      functional\n",
       "1  51630      functional\n",
       "2  17168      functional\n",
       "3  45559  non functional\n",
       "4  49871      functional"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ans_nn = pd.DataFrame({'id': df_test['id'], 'status_group': y_encoder.inverse_transform(y_te)})\n",
    "ans_nn.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "ans_nn.to_csv('ans_lightgbm.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Опробуем catboost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 265,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from catboost import Pool, CatBoostClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 266,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.51146464646464651, 0.0013627007639629043)"
      ]
     },
     "execution_count": 266,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf3 = CatBoostClassifier(random_seed=SEED, iterations=500, learning_rate=0.03, depth=6)\n",
    "\n",
    "scores = cross_val_score(clf3, x_train, y, n_jobs=-1)\n",
    "np.mean(scores), 2*np.std(scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 267,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "clf3 = clf3.fit(x_train, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 286,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.,  0.,  0., ...,  0.,  0.,  1.])"
      ]
     },
     "execution_count": 286,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_te = clf3.predict(x_test)\n",
    "y_te"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 290,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "arr = []\n",
    "\n",
    "i = 0\n",
    "while i < len(y_te):\n",
    "    arr.append(int(y_te[i]))\n",
    "    i += 1\n",
    "    \n",
    "#arr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 291,
   "metadata": {},
   "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>id</th>\n",
       "      <th>status_group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>50785</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>51630</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>17168</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>45559</td>\n",
       "      <td>functional needs repair</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>49871</td>\n",
       "      <td>functional</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id             status_group\n",
       "0  50785               functional\n",
       "1  51630               functional\n",
       "2  17168               functional\n",
       "3  45559  functional needs repair\n",
       "4  49871               functional"
      ]
     },
     "execution_count": 291,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ans_nn = pd.DataFrame({'id': df_test['id'], 'status_group': y_encoder.inverse_transform(arr)})\n",
    "ans_nn.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 292,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "ans_nn.to_csv('ans_catboost.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Опробуем H2O"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import h2o"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from h2o.estimators.gbm import H2OGradientBoostingEstimator\n",
    "from sklearn.model_selection import cross_val_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Checking whether there is an H2O instance running at http://localhost:54321..... not found.\n",
      "Attempting to start a local H2O server...\n",
      "; OpenJDK 64-Bit Server VM (Zulu 8.20.0.5-win64) (build 25.121-b15, mixed mode)\n",
      "  Starting server from C:\\ProgramData\\Anaconda3\\lib\\site-packages\\h2o\\backend\\bin\\h2o.jar\n",
      "  Ice root: C:\\Users\\Alice\\AppData\\Local\\Temp\\tmpbrdqc9kx\n",
      "  JVM stdout: C:\\Users\\Alice\\AppData\\Local\\Temp\\tmpbrdqc9kx\\h2o_Alice_started_from_python.out\n",
      "  JVM stderr: C:\\Users\\Alice\\AppData\\Local\\Temp\\tmpbrdqc9kx\\h2o_Alice_started_from_python.err\n",
      "  Server is running at http://127.0.0.1:54321\n",
      "Connecting to H2O server at http://127.0.0.1:54321... successful.\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td>H2O cluster uptime:</td>\n",
       "<td>06 secs</td></tr>\n",
       "<tr><td>H2O cluster version:</td>\n",
       "<td>3.10.4.8</td></tr>\n",
       "<tr><td>H2O cluster version age:</td>\n",
       "<td>2 months and 6 days </td></tr>\n",
       "<tr><td>H2O cluster name:</td>\n",
       "<td>H2O_from_python_Alice_vg1elo</td></tr>\n",
       "<tr><td>H2O cluster total nodes:</td>\n",
       "<td>1</td></tr>\n",
       "<tr><td>H2O cluster free memory:</td>\n",
       "<td>1.778 Gb</td></tr>\n",
       "<tr><td>H2O cluster total cores:</td>\n",
       "<td>4</td></tr>\n",
       "<tr><td>H2O cluster allowed cores:</td>\n",
       "<td>4</td></tr>\n",
       "<tr><td>H2O cluster status:</td>\n",
       "<td>accepting new members, healthy</td></tr>\n",
       "<tr><td>H2O connection url:</td>\n",
       "<td>http://127.0.0.1:54321</td></tr>\n",
       "<tr><td>H2O connection proxy:</td>\n",
       "<td>None</td></tr>\n",
       "<tr><td>H2O internal security:</td>\n",
       "<td>False</td></tr>\n",
       "<tr><td>Python version:</td>\n",
       "<td>3.6.1 final</td></tr></table></div>"
      ],
      "text/plain": [
       "--------------------------  ------------------------------\n",
       "H2O cluster uptime:         06 secs\n",
       "H2O cluster version:        3.10.4.8\n",
       "H2O cluster version age:    2 months and 6 days\n",
       "H2O cluster name:           H2O_from_python_Alice_vg1elo\n",
       "H2O cluster total nodes:    1\n",
       "H2O cluster free memory:    1.778 Gb\n",
       "H2O cluster total cores:    4\n",
       "H2O cluster allowed cores:  4\n",
       "H2O cluster status:         accepting new members, healthy\n",
       "H2O connection url:         http://127.0.0.1:54321\n",
       "H2O connection proxy:\n",
       "H2O internal security:      False\n",
       "Python version:             3.6.1 final\n",
       "--------------------------  ------------------------------"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "h2o.init()"
   ]
  },
  {
   "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": "3.6.1"
  },
  "toc": {
   "colors": {
    "hover_highlight": "#DAA520",
    "running_highlight": "#FF0000",
    "selected_highlight": "#FFD700"
   },
   "moveMenuLeft": true,
   "nav_menu": {
    "height": "174px",
    "width": "252px"
   },
   "navigate_menu": true,
   "number_sections": true,
   "sideBar": true,
   "threshold": 4,
   "toc_cell": false,
   "toc_section_display": "block",
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}