Trees.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import xgboost as xgb\n",
    "# import lightgbm as lgb\n",
    "import pickle\n",
    "import joblib\n",
    "\n",
    "from sklearn import tree\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.metrics import f1_score\n",
    "from xgboost import plot_tree\n",
    "\n",
    "from sklearn.tree import export_graphviz\n",
    "from IPython.display import SVG\n",
    "from graphviz import Source\n",
    "from IPython.display import display"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"data/one_hot_data.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>building_id</th>\n",
       "      <th>geo_level_1_id</th>\n",
       "      <th>geo_level_2_id</th>\n",
       "      <th>geo_level_3_id</th>\n",
       "      <th>count_floors_pre_eq</th>\n",
       "      <th>age</th>\n",
       "      <th>area_percentage</th>\n",
       "      <th>height_percentage</th>\n",
       "      <th>has_superstructure_adobe_mud</th>\n",
       "      <th>...</th>\n",
       "      <th>plan_configuration_m</th>\n",
       "      <th>plan_configuration_n</th>\n",
       "      <th>plan_configuration_o</th>\n",
       "      <th>plan_configuration_q</th>\n",
       "      <th>plan_configuration_s</th>\n",
       "      <th>plan_configuration_u</th>\n",
       "      <th>legal_ownership_status_a</th>\n",
       "      <th>legal_ownership_status_r</th>\n",
       "      <th>legal_ownership_status_v</th>\n",
       "      <th>legal_ownership_status_w</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>802906</td>\n",
       "      <td>6</td>\n",
       "      <td>487</td>\n",
       "      <td>12198</td>\n",
       "      <td>2</td>\n",
       "      <td>30</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>28830</td>\n",
       "      <td>8</td>\n",
       "      <td>900</td>\n",
       "      <td>2812</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>8</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>94947</td>\n",
       "      <td>21</td>\n",
       "      <td>363</td>\n",
       "      <td>8973</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>590882</td>\n",
       "      <td>22</td>\n",
       "      <td>418</td>\n",
       "      <td>10694</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>201944</td>\n",
       "      <td>11</td>\n",
       "      <td>131</td>\n",
       "      <td>1488</td>\n",
       "      <td>3</td>\n",
       "      <td>30</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 70 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0  building_id  geo_level_1_id  geo_level_2_id  geo_level_3_id  \\\n",
       "0           0       802906               6             487           12198   \n",
       "1           1        28830               8             900            2812   \n",
       "2           2        94947              21             363            8973   \n",
       "3           3       590882              22             418           10694   \n",
       "4           4       201944              11             131            1488   \n",
       "\n",
       "   count_floors_pre_eq  age  area_percentage  height_percentage  \\\n",
       "0                    2   30                6                  5   \n",
       "1                    2   10                8                  7   \n",
       "2                    2   10                5                  5   \n",
       "3                    2   10                6                  5   \n",
       "4                    3   30                8                  9   \n",
       "\n",
       "   has_superstructure_adobe_mud  ...  plan_configuration_m  \\\n",
       "0                             1  ...                     0   \n",
       "1                             0  ...                     0   \n",
       "2                             0  ...                     0   \n",
       "3                             0  ...                     0   \n",
       "4                             1  ...                     0   \n",
       "\n",
       "   plan_configuration_n  plan_configuration_o  plan_configuration_q  \\\n",
       "0                     0                     0                     0   \n",
       "1                     0                     0                     0   \n",
       "2                     0                     0                     0   \n",
       "3                     0                     0                     0   \n",
       "4                     0                     0                     0   \n",
       "\n",
       "   plan_configuration_s  plan_configuration_u  legal_ownership_status_a  \\\n",
       "0                     0                     0                         0   \n",
       "1                     0                     0                         0   \n",
       "2                     0                     0                         0   \n",
       "3                     0                     0                         0   \n",
       "4                     0                     0                         0   \n",
       "\n",
       "   legal_ownership_status_r  legal_ownership_status_v  \\\n",
       "0                         0                         1   \n",
       "1                         0                         1   \n",
       "2                         0                         1   \n",
       "3                         0                         1   \n",
       "4                         0                         1   \n",
       "\n",
       "   legal_ownership_status_w  \n",
       "0                         0  \n",
       "1                         0  \n",
       "2                         0  \n",
       "3                         0  \n",
       "4                         0  \n",
       "\n",
       "[5 rows x 70 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "df2 = pd.read_csv(\"data/train_labels.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>building_id</th>\n",
       "      <th>damage_grade</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>802906</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>28830</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>94947</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>590882</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>201944</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   building_id  damage_grade\n",
       "0       802906             3\n",
       "1        28830             2\n",
       "2        94947             3\n",
       "3       590882             2\n",
       "4       201944             3"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "test = df['building_id']\n",
    "test2 = df2['building_id']\n",
    "for i in range(len(test)):\n",
    "    if test[i] != test2[i]:\n",
    "        print(i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df.drop(['Unnamed: 0', 'building_id'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "temp = df.drop(['Unnamed: 0', 'building_id'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>geo_level_1_id</th>\n",
       "      <th>geo_level_2_id</th>\n",
       "      <th>geo_level_3_id</th>\n",
       "      <th>count_floors_pre_eq</th>\n",
       "      <th>age</th>\n",
       "      <th>area_percentage</th>\n",
       "      <th>height_percentage</th>\n",
       "      <th>has_superstructure_adobe_mud</th>\n",
       "      <th>has_superstructure_mud_mortar_stone</th>\n",
       "      <th>has_superstructure_stone_flag</th>\n",
       "      <th>...</th>\n",
       "      <th>plan_configuration_m</th>\n",
       "      <th>plan_configuration_n</th>\n",
       "      <th>plan_configuration_o</th>\n",
       "      <th>plan_configuration_q</th>\n",
       "      <th>plan_configuration_s</th>\n",
       "      <th>plan_configuration_u</th>\n",
       "      <th>legal_ownership_status_a</th>\n",
       "      <th>legal_ownership_status_r</th>\n",
       "      <th>legal_ownership_status_v</th>\n",
       "      <th>legal_ownership_status_w</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6</td>\n",
       "      <td>487</td>\n",
       "      <td>12198</td>\n",
       "      <td>2</td>\n",
       "      <td>30</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8</td>\n",
       "      <td>900</td>\n",
       "      <td>2812</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>8</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>21</td>\n",
       "      <td>363</td>\n",
       "      <td>8973</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>22</td>\n",
       "      <td>418</td>\n",
       "      <td>10694</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>11</td>\n",
       "      <td>131</td>\n",
       "      <td>1488</td>\n",
       "      <td>3</td>\n",
       "      <td>30</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 68 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   geo_level_1_id  geo_level_2_id  geo_level_3_id  count_floors_pre_eq  age  \\\n",
       "0               6             487           12198                    2   30   \n",
       "1               8             900            2812                    2   10   \n",
       "2              21             363            8973                    2   10   \n",
       "3              22             418           10694                    2   10   \n",
       "4              11             131            1488                    3   30   \n",
       "\n",
       "   area_percentage  height_percentage  has_superstructure_adobe_mud  \\\n",
       "0                6                  5                             1   \n",
       "1                8                  7                             0   \n",
       "2                5                  5                             0   \n",
       "3                6                  5                             0   \n",
       "4                8                  9                             1   \n",
       "\n",
       "   has_superstructure_mud_mortar_stone  has_superstructure_stone_flag  ...  \\\n",
       "0                                    1                              0  ...   \n",
       "1                                    1                              0  ...   \n",
       "2                                    1                              0  ...   \n",
       "3                                    1                              0  ...   \n",
       "4                                    0                              0  ...   \n",
       "\n",
       "   plan_configuration_m  plan_configuration_n  plan_configuration_o  \\\n",
       "0                     0                     0                     0   \n",
       "1                     0                     0                     0   \n",
       "2                     0                     0                     0   \n",
       "3                     0                     0                     0   \n",
       "4                     0                     0                     0   \n",
       "\n",
       "   plan_configuration_q  plan_configuration_s  plan_configuration_u  \\\n",
       "0                     0                     0                     0   \n",
       "1                     0                     0                     0   \n",
       "2                     0                     0                     0   \n",
       "3                     0                     0                     0   \n",
       "4                     0                     0                     0   \n",
       "\n",
       "   legal_ownership_status_a  legal_ownership_status_r  \\\n",
       "0                         0                         0   \n",
       "1                         0                         0   \n",
       "2                         0                         0   \n",
       "3                         0                         0   \n",
       "4                         0                         0   \n",
       "\n",
       "   legal_ownership_status_v  legal_ownership_status_w  \n",
       "0                         1                         0  \n",
       "1                         1                         0  \n",
       "2                         1                         0  \n",
       "3                         1                         0  \n",
       "4                         1                         0  \n",
       "\n",
       "[5 rows x 68 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "temp.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>geo_level_1_id</th>\n",
       "      <th>geo_level_2_id</th>\n",
       "      <th>geo_level_3_id</th>\n",
       "      <th>count_floors_pre_eq</th>\n",
       "      <th>age</th>\n",
       "      <th>area_percentage</th>\n",
       "      <th>height_percentage</th>\n",
       "      <th>has_superstructure_adobe_mud</th>\n",
       "      <th>has_superstructure_mud_mortar_stone</th>\n",
       "      <th>has_superstructure_stone_flag</th>\n",
       "      <th>...</th>\n",
       "      <th>plan_configuration_m</th>\n",
       "      <th>plan_configuration_n</th>\n",
       "      <th>plan_configuration_o</th>\n",
       "      <th>plan_configuration_q</th>\n",
       "      <th>plan_configuration_s</th>\n",
       "      <th>plan_configuration_u</th>\n",
       "      <th>legal_ownership_status_a</th>\n",
       "      <th>legal_ownership_status_r</th>\n",
       "      <th>legal_ownership_status_v</th>\n",
       "      <th>legal_ownership_status_w</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6</td>\n",
       "      <td>487</td>\n",
       "      <td>12198</td>\n",
       "      <td>2</td>\n",
       "      <td>30</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8</td>\n",
       "      <td>900</td>\n",
       "      <td>2812</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>8</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>21</td>\n",
       "      <td>363</td>\n",
       "      <td>8973</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>22</td>\n",
       "      <td>418</td>\n",
       "      <td>10694</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>11</td>\n",
       "      <td>131</td>\n",
       "      <td>1488</td>\n",
       "      <td>3</td>\n",
       "      <td>30</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 68 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   geo_level_1_id  geo_level_2_id  geo_level_3_id  count_floors_pre_eq  age  \\\n",
       "0               6             487           12198                    2   30   \n",
       "1               8             900            2812                    2   10   \n",
       "2              21             363            8973                    2   10   \n",
       "3              22             418           10694                    2   10   \n",
       "4              11             131            1488                    3   30   \n",
       "\n",
       "   area_percentage  height_percentage  has_superstructure_adobe_mud  \\\n",
       "0                6                  5                             1   \n",
       "1                8                  7                             0   \n",
       "2                5                  5                             0   \n",
       "3                6                  5                             0   \n",
       "4                8                  9                             1   \n",
       "\n",
       "   has_superstructure_mud_mortar_stone  has_superstructure_stone_flag  ...  \\\n",
       "0                                    1                              0  ...   \n",
       "1                                    1                              0  ...   \n",
       "2                                    1                              0  ...   \n",
       "3                                    1                              0  ...   \n",
       "4                                    0                              0  ...   \n",
       "\n",
       "   plan_configuration_m  plan_configuration_n  plan_configuration_o  \\\n",
       "0                     0                     0                     0   \n",
       "1                     0                     0                     0   \n",
       "2                     0                     0                     0   \n",
       "3                     0                     0                     0   \n",
       "4                     0                     0                     0   \n",
       "\n",
       "   plan_configuration_q  plan_configuration_s  plan_configuration_u  \\\n",
       "0                     0                     0                     0   \n",
       "1                     0                     0                     0   \n",
       "2                     0                     0                     0   \n",
       "3                     0                     0                     0   \n",
       "4                     0                     0                     0   \n",
       "\n",
       "   legal_ownership_status_a  legal_ownership_status_r  \\\n",
       "0                         0                         0   \n",
       "1                         0                         0   \n",
       "2                         0                         0   \n",
       "3                         0                         0   \n",
       "4                         0                         0   \n",
       "\n",
       "   legal_ownership_status_v  legal_ownership_status_w  \n",
       "0                         1                         0  \n",
       "1                         1                         0  \n",
       "2                         1                         0  \n",
       "3                         1                         0  \n",
       "4                         1                         0  \n",
       "\n",
       "[5 rows x 68 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "columns = temp.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = X.to_numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "Y = df2['damage_grade'].to_numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "clf = tree.DecisionTreeClassifier()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.64920713, 0.6503101 , 0.65186422])"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cross_val_score(clf, X_train, y_train, cv=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "clf = clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9867325402916348\n",
      "0.6579113984766217\n"
     ]
    }
   ],
   "source": [
    "# np.sum(clf.predict(X_test) == y_test) / len(df.index)\n",
    "print(f1_score(y_train, clf.predict(X_train), average='micro'))\n",
    "print(f1_score(y_test, clf.predict(X_test), average='micro'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "rf = RandomForestClassifier(n_estimators=3000, n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "# cross_val_score(rf, X_train, y_train, cv=3, n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "                       max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                       min_samples_leaf=1, min_samples_split=2,\n",
       "                       min_weight_fraction_leaf=0.0, n_estimators=3000,\n",
       "                       n_jobs=-1, oob_score=False, random_state=None, verbose=0,\n",
       "                       warm_start=False)"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = rf.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9867757099002302\n",
      "0.7137813932963681\n"
     ]
    }
   ],
   "source": [
    "# np.sum(predictions==y_test) / len(df.index)\n",
    "print(f1_score(y_train, rf.predict(X_train), average='micro'))\n",
    "print(f1_score(y_test, predictions, average='micro'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "# pickle.dump(rf, open(\"decision_models/rf3000.model\", 'wb'))\n",
    "joblib.dump(rf, \"decision_models/rf3000.model\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "gb = xgb.XGBClassifier(n_estimators=3000, n_jobs=-1)\n",
    "# parameters = {\n",
    "#      \"eta\"    : [0.05, 0.10, 0.15, 0.20, 0.25, 0.30 ] ,\n",
    "#      \"max_depth\"        : [ 3, 4, 5, 6, 8, 10, 12, 15, 30, 100],\n",
    "#      \"min_child_weight\" : [ 1, 3, 5, 7 ],\n",
    "#      \"gamma\"            : [ 0.0, 0.1, 0.2 , 0.3, 0.4 ],\n",
    "#      \"colsample_bytree\" : [ 0.3, 0.4, 0.5 , 0.7 ]\n",
    "# }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# grid = GridSearchCV(clf,\n",
    "#                     parameters, n_jobs=4,\n",
    "#                     scoring=\"neg_log_loss\",\n",
    "#                     cv=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "XGBClassifier(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
       "              colsample_bytree=1, gamma=0, learning_rate=0.1, max_delta_step=0,\n",
       "              max_depth=3, min_child_weight=1, missing=None, n_estimators=3000,\n",
       "              n_jobs=-1, nthread=None, objective='multi:softprob',\n",
       "              random_state=0, reg_alpha=0, reg_lambda=1, scale_pos_weight=1,\n",
       "              seed=None, silent=True, subsample=1)"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gb.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "gb_predictions = gb.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7507818495778972\n",
      "0.7365361370656741\n"
     ]
    }
   ],
   "source": [
    "print(f1_score(y_train, gb.predict(X_train), average='micro'))\n",
    "print(f1_score(y_test, gb_predictions, average='micro'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [],
   "source": [
    "gbf = xgb.XGBClassifier(n_estimators=3000, objective='multi:softmax', num_class=3, n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "XGBClassifier(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
       "              colsample_bytree=1, gamma=0, learning_rate=0.1, max_delta_step=0,\n",
       "              max_depth=3, min_child_weight=1, missing=None, n_estimators=3000,\n",
       "              n_jobs=-1, nthread=None, num_class=3, objective='multi:softprob',\n",
       "              random_state=0, reg_alpha=0, reg_lambda=1, scale_pos_weight=1,\n",
       "              seed=None, silent=True, subsample=1)"
      ]
     },
     "execution_count": 144,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gbf.fit(X, Y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 1080x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# feat_imp = pd.Series(gbf.booster().get_fscore()).sort_values(ascending=False)\n",
    "# feat_imp.plot(kind='bar', title='Feature Importances')\n",
    "# plt.ylabel('Feature Importance Score')\n",
    "# plt.savefig(\"xgboostfeatimp.png\")\n",
    "\n",
    "# importances = gbf.feature_importances_\n",
    "# std = np.std([tree.feature_importances_ for tree in gbf.estimators_],\n",
    "#              axis=0)\n",
    "# indices = np.argsort(importances)[::-1]\n",
    "\n",
    "# # Print the feature ranking\n",
    "# print(\"Feature ranking:\")\n",
    "\n",
    "# for f in range(X.shape[1]):\n",
    "#     print(\"%d. feature %d (%f)\" % (f + 1, indices[f], importances[indices[f]]))\n",
    "\n",
    "# # Plot the feature importances of the forest\n",
    "# plt.figure()\n",
    "# plt.title(\"Feature importances\")\n",
    "# plt.bar(range(X.shape[1]), importances[indices],\n",
    "#        color=\"r\", yerr=std[indices], align=\"center\")\n",
    "# plt.xticks(range(X.shape[1]), indices)\n",
    "# plt.xlim([-1, X.shape[1]])\n",
    "# plt.show()\n",
    "plt.figure(figsize=(15,4))\n",
    "xgb.plot_importance(gbf)\n",
    "plt.show()\n",
    "plt.savefig(\"f_importance_xgboostf3000\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7498369148119723\n",
      "0.7508297998887205\n"
     ]
    }
   ],
   "source": [
    "print(f1_score(y_train, gbf.predict(X_train), average='micro'))\n",
    "print(f1_score(y_test, gbf.predict(X_test), average='micro'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['decision_models/xgbf3000.model']"
      ]
     },
     "execution_count": 175,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "joblib.dump(gbf, \"decision_models/xgbf3000.model\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_tree(gb, num_trees=4)\n",
    "# tree.plot_tree(gbf)\n",
    "plt.savefig(\"xgboostf4.png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "test = pd.read_csv('data/test_values.csv', index_col='building_id')\n",
    "test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "def oneHot(df, column_name):\n",
    "    one_hot = pd.get_dummies(df[column_name])\n",
    "    df = df.drop(column_name,axis = 1)\n",
    "    df = df.join(one_hot)\n",
    "    for letter in one_hot.columns:\n",
    "        df[column_name + \"_\" + letter] = df[letter]\n",
    "        df.drop(letter, axis=1, inplace=True)\n",
    "    return df\n",
    "\n",
    "categorical_vars = [\"foundation_type\", \"land_surface_condition\", \"roof_type\", \"ground_floor_type\", \"other_floor_type\", \"position\", \"plan_configuration\", \"legal_ownership_status\"]\n",
    "for var in categorical_vars:\n",
    "    test = oneHot(test, var)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "68 68\n"
     ]
    }
   ],
   "source": [
    "print(len(test.columns), len(columns))\n",
    "\n",
    "for i in range (len(test.columns)):\n",
    "\n",
    "    if test.columns[i] != columns[i]:\n",
    "        print(test.columns[i], columns[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [],
   "source": [
    "test = test.reindex(columns=columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [],
   "source": [
    "test1 = test.to_numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_pred = gbf.predict(test1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [],
   "source": [
    "submission_format = pd.read_csv('data/submission_format.csv', index_col='building_id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [],
   "source": [
    "my_submission = pd.DataFrame(data=test_pred,\n",
    "                             columns=submission_format.columns,\n",
    "                             index=submission_format.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>damage_grade</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>building_id</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>300051</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99355</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>890251</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>745817</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>421793</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             damage_grade\n",
       "building_id              \n",
       "300051                  3\n",
       "99355                   2\n",
       "890251                  2\n",
       "745817                  1\n",
       "421793                  3"
      ]
     },
     "execution_count": 159,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "my_submission.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {},
   "outputs": [],
   "source": [
    "my_submission.to_csv('data/submission1.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "joblib.dump(gb, \"decision_models/gb3000.model\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "ngb = xgb.XGBClassifier(n_estimators=1000, n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "XGBClassifier(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
       "              colsample_bytree=1, gamma=0, learning_rate=0.1, max_delta_step=0,\n",
       "              max_depth=3, min_child_weight=1, missing=None, n_estimators=1000,\n",
       "              n_jobs=-1, nthread=None, objective='multi:softprob',\n",
       "              random_state=0, reg_alpha=0, reg_lambda=1, scale_pos_weight=1,\n",
       "              seed=None, silent=True, subsample=1)"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ngb.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7261655794320797\n",
      "0.7211872373899196\n"
     ]
    }
   ],
   "source": [
    "print(f1_score(y_train, ngb.predict(X_train), average='micro'))\n",
    "print(f1_score(y_test, ngb.predict(X_test), average='micro'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "opt_xgb = xgb.XGBClassifier()\n",
    "parameters = {\n",
    "     \"eta\"    : [0.05, 0.10, 0.15, 0.20, 0.25, 0.30 ] ,\n",
    "     \"max_depth\"        : [ 3, 4, 5, 6, 8, 10, 12, 15],\n",
    "     \"min_child_weight\" : [ 1, 3, 5, 7 ],\n",
    "     \"gamma\"            : [ 0.0, 0.1, 0.2 , 0.3, 0.4 ],\n",
    "     \"colsample_bytree\" : [ 0.3, 0.4, 0.5 , 0.7 ]\n",
    "     }\n",
    "\n",
    "grid = GridSearchCV(opt_xgb,\n",
    "                    parameters, n_jobs=-1,\n",
    "                    scoring=\"f1_micro\",\n",
    "                    cv=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "Invalid parameter eta for estimator XGBClassifier(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n              colsample_bytree=0.3, gamma=0, learning_rate=0.1,\n              max_delta_step=0, max_depth=3, min_child_weight=1, missing=nan,\n              n_estimators=100, n_jobs=1, nthread=None,\n              objective='binary:logistic', random_state=0, reg_alpha=0,\n              reg_lambda=1, scale_pos_weight=1, seed=None, silent=True,\n              subsample=1). Check the list of available parameters with `estimator.get_params().keys()`.",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31m_RemoteTraceback\u001b[0m                          Traceback (most recent call last)",
      "\u001b[0;31m_RemoteTraceback\u001b[0m: \n\"\"\"\nTraceback (most recent call last):\n  File \"/Users/naman/.pyenv/versions/3.6.8/lib/python3.6/site-packages/joblib/externals/loky/process_executor.py\", line 418, in _process_worker\n    r = call_item()\n  File \"/Users/naman/.pyenv/versions/3.6.8/lib/python3.6/site-packages/joblib/externals/loky/process_executor.py\", line 272, in __call__\n    return self.fn(*self.args, **self.kwargs)\n  File \"/Users/naman/.pyenv/versions/3.6.8/lib/python3.6/site-packages/joblib/_parallel_backends.py\", line 600, in __call__\n    return self.func(*args, **kwargs)\n  File \"/Users/naman/.pyenv/versions/3.6.8/lib/python3.6/site-packages/joblib/parallel.py\", line 256, in __call__\n    for func, args, kwargs in self.items]\n  File \"/Users/naman/.pyenv/versions/3.6.8/lib/python3.6/site-packages/joblib/parallel.py\", line 256, in <listcomp>\n    for func, args, kwargs in self.items]\n  File \"/Users/naman/.pyenv/versions/3.6.8/lib/python3.6/site-packages/sklearn/model_selection/_validation.py\", line 503, in _fit_and_score\n    estimator.set_params(**parameters)\n  File \"/Users/naman/.pyenv/versions/3.6.8/lib/python3.6/site-packages/sklearn/base.py\", line 224, in set_params\n    (key, self))\nValueError: Invalid parameter eta for estimator XGBClassifier(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n              colsample_bytree=0.3, gamma=0, learning_rate=0.1,\n              max_delta_step=0, max_depth=3, min_child_weight=1, missing=nan,\n              n_estimators=100, n_jobs=1, nthread=None,\n              objective='binary:logistic', random_state=0, reg_alpha=0,\n              reg_lambda=1, scale_pos_weight=1, seed=None, silent=True,\n              subsample=1). Check the list of available parameters with `estimator.get_params().keys()`.\n\"\"\"",
      "\nThe above exception was the direct cause of the following exception:\n",
      "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-63-3949096c802a>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mgrid\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX_train\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_train\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m~/.pyenv/versions/3.6.8/lib/python3.6/site-packages/sklearn/model_selection/_search.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, X, y, groups, **fit_params)\u001b[0m\n\u001b[1;32m    686\u001b[0m                 \u001b[0;32mreturn\u001b[0m \u001b[0mresults\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    687\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 688\u001b[0;31m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_run_search\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mevaluate_candidates\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    689\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    690\u001b[0m         \u001b[0;31m# For multi-metric evaluation, store the best_index_, best_params_ and\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.6.8/lib/python3.6/site-packages/sklearn/model_selection/_search.py\u001b[0m in \u001b[0;36m_run_search\u001b[0;34m(self, evaluate_candidates)\u001b[0m\n\u001b[1;32m   1147\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_run_search\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mevaluate_candidates\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1148\u001b[0m         \u001b[0;34m\"\"\"Search all candidates in param_grid\"\"\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1149\u001b[0;31m         \u001b[0mevaluate_candidates\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mParameterGrid\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mparam_grid\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1150\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1151\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.6.8/lib/python3.6/site-packages/sklearn/model_selection/_search.py\u001b[0m in \u001b[0;36mevaluate_candidates\u001b[0;34m(candidate_params)\u001b[0m\n\u001b[1;32m    665\u001b[0m                                \u001b[0;32mfor\u001b[0m \u001b[0mparameters\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mtrain\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtest\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    666\u001b[0m                                in product(candidate_params,\n\u001b[0;32m--> 667\u001b[0;31m                                           cv.split(X, y, groups)))\n\u001b[0m\u001b[1;32m    668\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    669\u001b[0m                 \u001b[0;32mif\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mout\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m<\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.6.8/lib/python3.6/site-packages/joblib/parallel.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, iterable)\u001b[0m\n\u001b[1;32m   1014\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1015\u001b[0m             \u001b[0;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_backend\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mretrieval_context\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1016\u001b[0;31m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mretrieve\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1017\u001b[0m             \u001b[0;31m# Make sure that we get a last message telling us we are done\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1018\u001b[0m             \u001b[0melapsed_time\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtime\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtime\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m-\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_start_time\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.6.8/lib/python3.6/site-packages/joblib/parallel.py\u001b[0m in \u001b[0;36mretrieve\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    906\u001b[0m             \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    907\u001b[0m                 \u001b[0;32mif\u001b[0m \u001b[0mgetattr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_backend\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'supports_timeout'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 908\u001b[0;31m                     \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_output\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mextend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mjob\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtimeout\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtimeout\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    909\u001b[0m                 \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    910\u001b[0m                     \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_output\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mextend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mjob\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.6.8/lib/python3.6/site-packages/joblib/_parallel_backends.py\u001b[0m in \u001b[0;36mwrap_future_result\u001b[0;34m(future, timeout)\u001b[0m\n\u001b[1;32m    552\u001b[0m         AsyncResults.get from multiprocessing.\"\"\"\n\u001b[1;32m    553\u001b[0m         \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 554\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mfuture\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mresult\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtimeout\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtimeout\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    555\u001b[0m         \u001b[0;32mexcept\u001b[0m \u001b[0mLokyTimeoutError\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    556\u001b[0m             \u001b[0;32mraise\u001b[0m \u001b[0mTimeoutError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.6.8/lib/python3.6/concurrent/futures/_base.py\u001b[0m in \u001b[0;36mresult\u001b[0;34m(self, timeout)\u001b[0m\n\u001b[1;32m    430\u001b[0m                 \u001b[0;32mraise\u001b[0m \u001b[0mCancelledError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    431\u001b[0m             \u001b[0;32melif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_state\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0mFINISHED\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 432\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__get_result\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    433\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    434\u001b[0m                 \u001b[0;32mraise\u001b[0m \u001b[0mTimeoutError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.6.8/lib/python3.6/concurrent/futures/_base.py\u001b[0m in \u001b[0;36m__get_result\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    382\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m__get_result\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    383\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_exception\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 384\u001b[0;31m             \u001b[0;32mraise\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_exception\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    385\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    386\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_result\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mValueError\u001b[0m: Invalid parameter eta for estimator XGBClassifier(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n              colsample_bytree=0.3, gamma=0, learning_rate=0.1,\n              max_delta_step=0, max_depth=3, min_child_weight=1, missing=nan,\n              n_estimators=100, n_jobs=1, nthread=None,\n              objective='binary:logistic', random_state=0, reg_alpha=0,\n              reg_lambda=1, scale_pos_weight=1, seed=None, silent=True,\n              subsample=1). Check the list of available parameters with `estimator.get_params().keys()`."
     ]
    }
   ],
   "source": [
    "grid.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print(f1_score(y_train, grid.predict(X_train), average='micro'))\n",
    "print(f1_score(y_test, grid.predict(X_test), average='micro'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d_train = lgb.Dataset(X_train, label=y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "xs = np.arange(4)\n",
    "# F1_baselines = [0.6549, 0.7137, 0.7365, 0.7226]\n",
    "F1_baselines_train = [0.9867, 0.9867, 0.7542, 0.7823]\n",
    "F1_baselines_test = [0.6549, 0.6911, 0.7046, 0.7119]\n",
    "F1_baseline_labels = ['Decision Tree', 'Random Forest', 'Extreme Gradient Boosting', 'Lite Gradient Boosting']\n",
    "plt.figure(figsize=(11,6))\n",
    "plt.bar(xs-0.375, F1_baselines_train, align='center', color='tab:blue', label='Train', width=0.375)\n",
    "plt.bar(xs, F1_baselines_test, align='center', color='tab:orange', label='Validation', width=0.375)\n",
    "plt.legend()\n",
    "plt.xticks(xs, F1_baseline_labels)\n",
    "plt.xlabel('Tree Type')\n",
    "plt.ylabel('F1 Score')\n",
    "plt.title('Baseline Tree Type F1 Score Comparison')\n",
    "plt.savefig('source_images/baseline_tree_comp.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "xs = np.arange(4)\n",
    "F1_baselines_train = [0.9867, 0.9867, 0.7542, 0.7823]\n",
    "F1_baselines_test = [0.6549, 0.6911, 0.7046, 0.7119]\n",
    "F1_final_train = [0.9867, 0.9867, 0.7498, 0.8194]\n",
    "F1_final_test = [0.6549, 0.7137, 0.7376, 0.7462]\n",
    "F1_baseline_labels = ['Decision Tree', 'Random Forest', 'Extreme Gradient Boosting', 'Lite Gradient Boosting']\n",
    "plt.figure(figsize=(12,6))\n",
    "plt.bar(xs-0.45, F1_baselines_train, align='center', color='tab:blue', label='Train', width=0.15)\n",
    "plt.bar(xs-0.3, F1_baselines_test, align='center', color='tab:orange', label='Validation', width=0.15)\n",
    "plt.bar(xs-0.15, F1_final_train, align='center', color='tab:red', label='Tuned Train', width=0.15)\n",
    "plt.bar(xs, F1_final_test, align='center', color='tab:green', label='Tuned Validation', width=0.15)\n",
    "plt.legend(loc=\"lower left\")\n",
    "plt.xticks(xs, F1_baseline_labels)\n",
    "plt.xlabel('Tree Type')\n",
    "plt.ylabel('F1 Score')\n",
    "plt.title('Tree Type F1 Score Comparison')\n",
    "plt.savefig('source_images/tuned_tree_comp.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "xs3000 = [30, 31, 32, 33, 34, 35, 40, 50, 60, 100]\n",
    "ys3000 = [0.7470, 0.7475, 0.7482, 0.7470, 0.7479, 0.7455, 0.7453, 0.7468, 0.7470, 0.743]\n",
    "plt.figure(figsize=(12,6))\n",
    "plt.plot(xs3000, ys3000, '-ok');\n",
    "plt.xlabel('Number of Leaves')\n",
    "plt.ylabel('F1 Score')\n",
    "plt.title('Lite Gradient Boosting Ensemble Validation accuracy for Number of Leaves')\n",
    "plt.savefig('source_images/num_leaves_analysis.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "3.6.8",
   "language": "python",
   "name": "3.6.8"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}