cv_load_Xgboost.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from collections import Counter\n",
    "\n",
    "dtype = {\n",
    "    'Usage': 'category',\n",
    "    'Description': 'category',\n",
    "    'status': 'category',\n",
    "}\n",
    "df = pd.read_csv('data/reservations.csv.gz', dtype=dtype, parse_dates=['created', 'arrival', 'departure'])\n",
    "df.drop(columns='Usage', inplace=True)\n",
    "\n",
    "df.loc[df['cancel_date'] == '0001-01-01T00:00:00', ['cancel_date']] = None\n",
    "df['cancel_date'] = pd.to_datetime(df['cancel_date'])\n",
    "\n",
    "df['arrival_year'] = df['arrival'].dt.year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "appearances = {}\n",
    "for contract_id in df['contract_id'].unique():\n",
    "    subset_df = df.loc[df['contract_id'] == contract_id].sort_values(by=['arrival', 'created'])\n",
    "    # save the last known state\n",
    "    # fill in blanks for bad years\n",
    "    # handle case where cancel year might come after a series of misses\n",
    "    yearly_state = {arrival_year: status for (created, arrival_year, status) in subset_df[['created', 'arrival_year', 'status']].itertuples(index=False, name=None)}\n",
    "    earliest = subset_df['arrival_year'].min()\n",
    "    latest = min(subset_df['arrival_year'].max(), 2019)\n",
    "    activity = [(year, yearly_state.get(year, 'no-show')) for year in range(earliest, latest + 1)]    \n",
    "    if activity:\n",
    "        resort_id = subset_df['resort_id'].values[0]\n",
    "        appearances[str(contract_id)] = [resort_id] + activity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "rows = []\n",
    "for r in list(appearances.values()):\n",
    "    resort_id, activity = r[0], r[1:]\n",
    "    row = [None] * 5\n",
    "    row[-len(activity):] = [s for year, s in activity]\n",
    "    rows.append([resort_id] + row)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.DataFrame(rows, columns=['resort_id', 'year_2015', 'year_2016', 'year_2017', 'year_2018', 'year_2019'])"
   ]
  },
  {
   "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>resort_id</th>\n",
       "      <th>year_2015</th>\n",
       "      <th>year_2016</th>\n",
       "      <th>year_2017</th>\n",
       "      <th>year_2018</th>\n",
       "      <th>year_2019</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>28</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>active</td>\n",
       "      <td>active</td>\n",
       "      <td>active</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>28</td>\n",
       "      <td>active</td>\n",
       "      <td>active</td>\n",
       "      <td>active</td>\n",
       "      <td>active</td>\n",
       "      <td>active</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>30</td>\n",
       "      <td>active</td>\n",
       "      <td>active</td>\n",
       "      <td>active</td>\n",
       "      <td>cancelled</td>\n",
       "      <td>cancelled</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>39</td>\n",
       "      <td>no-show</td>\n",
       "      <td>active</td>\n",
       "      <td>active</td>\n",
       "      <td>active</td>\n",
       "      <td>active</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>38</td>\n",
       "      <td>cancelled</td>\n",
       "      <td>active</td>\n",
       "      <td>cancelled</td>\n",
       "      <td>active</td>\n",
       "      <td>active</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   resort_id  year_2015 year_2016  year_2017  year_2018  year_2019\n",
       "0         28       None      None     active     active     active\n",
       "1         28     active    active     active     active     active\n",
       "2         30     active    active     active  cancelled  cancelled\n",
       "3         39    no-show    active     active     active     active\n",
       "4         38  cancelled    active  cancelled     active     active"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.drop('resort_id', axis=1, inplace=True)\n",
    "\n",
    "df.year_2015 = pd.Categorical(df.year_2015)\n",
    "df['year_2015'] = df.year_2015.cat.codes\n",
    "\n",
    "df.year_2016 = pd.Categorical(df.year_2016)\n",
    "df['year_2016'] = df.year_2016.cat.codes\n",
    "\n",
    "df.year_2017 = pd.Categorical(df.year_2017)\n",
    "df['year_2017'] = df.year_2017.cat.codes\n",
    "\n",
    "df.year_2018 = pd.Categorical(df.year_2018)\n",
    "df['year_2018'] = df.year_2018.cat.codes\n",
    "\n",
    "#df.year_2019 = pd.Categorical(df.year_2019)\n",
    "#df['year_2019'] = df.year_2019.cat.codes\n",
    "\n",
    "df.columns = [''] * len(df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "from xgboost import XGBClassifier\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import accuracy_score\n",
    "from xgboost import plot_importance\n",
    "from sklearn.metrics import roc_auc_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df.values[:,:-1]\n",
    "y = df.values[:, -1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.6675513946530258\n"
     ]
    }
   ],
   "source": [
    "model = XGBClassifier()\n",
    "model.fit(X_train, y_train)\n",
    "y_pred = model.predict(X_test)\n",
    "accuracy = accuracy_score(y_test, y_pred)\n",
    "\n",
    "print('Accuracy:', accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC: 0.7678660585317042\n"
     ]
    }
   ],
   "source": [
    "y_prob = model.predict_proba(X_test)\n",
    "auc = roc_auc_score(y_test, y_prob, multi_class=\"ovo\", average=\"macro\")\n",
    "\n",
    "print('AUC:', auc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "      active       0.71      0.86      0.78     11582\n",
      "   cancelled       0.61      0.51      0.55      5562\n",
      "     no-show       0.51      0.29      0.37      3578\n",
      "\n",
      "    accuracy                           0.67     20722\n",
      "   macro avg       0.61      0.55      0.57     20722\n",
      "weighted avg       0.65      0.67      0.65     20722\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "\n",
    "print(classification_report(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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": [
    "model.get_booster().feature_names = ['year_2015', 'year_2016', 'year_2017', 'year_2018']\n",
    "plot_importance(model.get_booster());"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0]\tvalidation_0-merror:0.3396\tvalidation_0-mlogloss:1.05477\tvalidation_1-merror:0.339301\tvalidation_1-mlogloss:1.05502\n",
      "[1]\tvalidation_0-merror:0.34142\tvalidation_0-mlogloss:1.01778\tvalidation_1-merror:0.34099\tvalidation_1-mlogloss:1.01841\n",
      "[2]\tvalidation_0-merror:0.34142\tvalidation_0-mlogloss:0.986663\tvalidation_1-merror:0.34099\tvalidation_1-mlogloss:0.987349\n",
      "[3]\tvalidation_0-merror:0.34142\tvalidation_0-mlogloss:0.960197\tvalidation_1-merror:0.34099\tvalidation_1-mlogloss:0.96085\n",
      "[4]\tvalidation_0-merror:0.34142\tvalidation_0-mlogloss:0.937143\tvalidation_1-merror:0.34099\tvalidation_1-mlogloss:0.938186\n",
      "[5]\tvalidation_0-merror:0.34142\tvalidation_0-mlogloss:0.917799\tvalidation_1-merror:0.34099\tvalidation_1-mlogloss:0.918738\n",
      "[6]\tvalidation_0-merror:0.34142\tvalidation_0-mlogloss:0.901054\tvalidation_1-merror:0.34099\tvalidation_1-mlogloss:0.901939\n",
      "[7]\tvalidation_0-merror:0.34142\tvalidation_0-mlogloss:0.8865\tvalidation_1-merror:0.34099\tvalidation_1-mlogloss:0.887359\n",
      "[8]\tvalidation_0-merror:0.34142\tvalidation_0-mlogloss:0.873882\tvalidation_1-merror:0.34099\tvalidation_1-mlogloss:0.87476\n",
      "[9]\tvalidation_0-merror:0.34142\tvalidation_0-mlogloss:0.863038\tvalidation_1-merror:0.34099\tvalidation_1-mlogloss:0.863981\n",
      "[10]\tvalidation_0-merror:0.33991\tvalidation_0-mlogloss:0.853589\tvalidation_1-merror:0.340266\tvalidation_1-mlogloss:0.854481\n",
      "[11]\tvalidation_0-merror:0.33991\tvalidation_0-mlogloss:0.845369\tvalidation_1-merror:0.340266\tvalidation_1-mlogloss:0.846237\n",
      "[12]\tvalidation_0-merror:0.33991\tvalidation_0-mlogloss:0.838393\tvalidation_1-merror:0.340266\tvalidation_1-mlogloss:0.839179\n",
      "[13]\tvalidation_0-merror:0.33991\tvalidation_0-mlogloss:0.832054\tvalidation_1-merror:0.340266\tvalidation_1-mlogloss:0.832772\n",
      "[14]\tvalidation_0-merror:0.33991\tvalidation_0-mlogloss:0.826918\tvalidation_1-merror:0.340266\tvalidation_1-mlogloss:0.827643\n",
      "[15]\tvalidation_0-merror:0.33991\tvalidation_0-mlogloss:0.82222\tvalidation_1-merror:0.340266\tvalidation_1-mlogloss:0.82291\n",
      "[16]\tvalidation_0-merror:0.33991\tvalidation_0-mlogloss:0.818291\tvalidation_1-merror:0.340266\tvalidation_1-mlogloss:0.818941\n",
      "[17]\tvalidation_0-merror:0.339455\tvalidation_0-mlogloss:0.814533\tvalidation_1-merror:0.340556\tvalidation_1-mlogloss:0.815285\n",
      "[18]\tvalidation_0-merror:0.339435\tvalidation_0-mlogloss:0.811274\tvalidation_1-merror:0.34017\tvalidation_1-mlogloss:0.812117\n",
      "[19]\tvalidation_0-merror:0.339435\tvalidation_0-mlogloss:0.808605\tvalidation_1-merror:0.34017\tvalidation_1-mlogloss:0.809398\n",
      "[20]\tvalidation_0-merror:0.335339\tvalidation_0-mlogloss:0.805982\tvalidation_1-merror:0.335199\tvalidation_1-mlogloss:0.806835\n",
      "[21]\tvalidation_0-merror:0.335339\tvalidation_0-mlogloss:0.803714\tvalidation_1-merror:0.335199\tvalidation_1-mlogloss:0.804678\n",
      "[22]\tvalidation_0-merror:0.334347\tvalidation_0-mlogloss:0.801768\tvalidation_1-merror:0.333945\tvalidation_1-mlogloss:0.802672\n",
      "[23]\tvalidation_0-merror:0.33414\tvalidation_0-mlogloss:0.800066\tvalidation_1-merror:0.334427\tvalidation_1-mlogloss:0.801081\n",
      "[24]\tvalidation_0-merror:0.33414\tvalidation_0-mlogloss:0.798531\tvalidation_1-merror:0.334427\tvalidation_1-mlogloss:0.799468\n",
      "[25]\tvalidation_0-merror:0.333437\tvalidation_0-mlogloss:0.796854\tvalidation_1-merror:0.333655\tvalidation_1-mlogloss:0.797836\n",
      "[26]\tvalidation_0-merror:0.333437\tvalidation_0-mlogloss:0.795511\tvalidation_1-merror:0.333655\tvalidation_1-mlogloss:0.796422\n",
      "[27]\tvalidation_0-merror:0.333437\tvalidation_0-mlogloss:0.794246\tvalidation_1-merror:0.333655\tvalidation_1-mlogloss:0.795084\n",
      "[28]\tvalidation_0-merror:0.333437\tvalidation_0-mlogloss:0.793096\tvalidation_1-merror:0.333655\tvalidation_1-mlogloss:0.794\n",
      "[29]\tvalidation_0-merror:0.333106\tvalidation_0-mlogloss:0.792061\tvalidation_1-merror:0.333607\tvalidation_1-mlogloss:0.792862\n",
      "[30]\tvalidation_0-merror:0.333561\tvalidation_0-mlogloss:0.791092\tvalidation_1-merror:0.333317\tvalidation_1-mlogloss:0.791882\n",
      "[31]\tvalidation_0-merror:0.333561\tvalidation_0-mlogloss:0.790295\tvalidation_1-merror:0.333317\tvalidation_1-mlogloss:0.791011\n",
      "[32]\tvalidation_0-merror:0.333106\tvalidation_0-mlogloss:0.789673\tvalidation_1-merror:0.333607\tvalidation_1-mlogloss:0.790319\n",
      "[33]\tvalidation_0-merror:0.333064\tvalidation_0-mlogloss:0.788942\tvalidation_1-merror:0.332449\tvalidation_1-mlogloss:0.789537\n",
      "[34]\tvalidation_0-merror:0.332547\tvalidation_0-mlogloss:0.788349\tvalidation_1-merror:0.332497\tvalidation_1-mlogloss:0.788995\n",
      "[35]\tvalidation_0-merror:0.333002\tvalidation_0-mlogloss:0.787743\tvalidation_1-merror:0.332207\tvalidation_1-mlogloss:0.7884\n",
      "[36]\tvalidation_0-merror:0.333002\tvalidation_0-mlogloss:0.787225\tvalidation_1-merror:0.332207\tvalidation_1-mlogloss:0.787923\n",
      "[37]\tvalidation_0-merror:0.333002\tvalidation_0-mlogloss:0.78691\tvalidation_1-merror:0.332207\tvalidation_1-mlogloss:0.78761\n",
      "[38]\tvalidation_0-merror:0.333002\tvalidation_0-mlogloss:0.786591\tvalidation_1-merror:0.332207\tvalidation_1-mlogloss:0.787204\n",
      "[39]\tvalidation_0-merror:0.333002\tvalidation_0-mlogloss:0.786228\tvalidation_1-merror:0.332207\tvalidation_1-mlogloss:0.786946\n",
      "[40]\tvalidation_0-merror:0.333002\tvalidation_0-mlogloss:0.785962\tvalidation_1-merror:0.332207\tvalidation_1-mlogloss:0.786482\n",
      "[41]\tvalidation_0-merror:0.333002\tvalidation_0-mlogloss:0.785637\tvalidation_1-merror:0.332207\tvalidation_1-mlogloss:0.786239\n",
      "[42]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.785407\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.785975\n",
      "[43]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.785101\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.785686\n",
      "[44]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.784929\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.785492\n",
      "[45]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.784765\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.785343\n",
      "[46]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.784674\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.7851\n",
      "[47]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.784411\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.784982\n",
      "[48]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.784272\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.784788\n",
      "[49]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.784274\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.784645\n",
      "[50]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.784081\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.784543\n",
      "[51]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.783898\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.784427\n",
      "[52]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.783846\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.784326\n",
      "[53]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.78386\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.784323\n",
      "[54]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.783729\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.784178\n",
      "[55]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.78358\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.784117\n",
      "[56]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.783509\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.784047\n",
      "[57]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.783505\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.784026\n",
      "[58]\tvalidation_0-merror:0.332506\tvalidation_0-mlogloss:0.783426\tvalidation_1-merror:0.332063\tvalidation_1-mlogloss:0.783906\n",
      "[59]\tvalidation_0-merror:0.332651\tvalidation_0-mlogloss:0.783269\tvalidation_1-merror:0.332497\tvalidation_1-mlogloss:0.783847\n",
      "[60]\tvalidation_0-merror:0.332651\tvalidation_0-mlogloss:0.783168\tvalidation_1-merror:0.332497\tvalidation_1-mlogloss:0.783816\n",
      "[61]\tvalidation_0-merror:0.332651\tvalidation_0-mlogloss:0.783262\tvalidation_1-merror:0.332497\tvalidation_1-mlogloss:0.783741\n",
      "[62]\tvalidation_0-merror:0.332651\tvalidation_0-mlogloss:0.783172\tvalidation_1-merror:0.332497\tvalidation_1-mlogloss:0.783703\n",
      "[63]\tvalidation_0-merror:0.332651\tvalidation_0-mlogloss:0.783041\tvalidation_1-merror:0.332497\tvalidation_1-mlogloss:0.783646\n",
      "[64]\tvalidation_0-merror:0.332651\tvalidation_0-mlogloss:0.782997\tvalidation_1-merror:0.332497\tvalidation_1-mlogloss:0.783636\n",
      "[65]\tvalidation_0-merror:0.332651\tvalidation_0-mlogloss:0.783023\tvalidation_1-merror:0.332497\tvalidation_1-mlogloss:0.783575\n",
      "[66]\tvalidation_0-merror:0.332651\tvalidation_0-mlogloss:0.782997\tvalidation_1-merror:0.332497\tvalidation_1-mlogloss:0.783515\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[67]\tvalidation_0-merror:0.332651\tvalidation_0-mlogloss:0.782833\tvalidation_1-merror:0.332497\tvalidation_1-mlogloss:0.783534\n",
      "[68]\tvalidation_0-merror:0.332651\tvalidation_0-mlogloss:0.782839\tvalidation_1-merror:0.332497\tvalidation_1-mlogloss:0.783478\n",
      "[69]\tvalidation_0-merror:0.332651\tvalidation_0-mlogloss:0.782821\tvalidation_1-merror:0.332497\tvalidation_1-mlogloss:0.783481\n",
      "[70]\tvalidation_0-merror:0.332651\tvalidation_0-mlogloss:0.782844\tvalidation_1-merror:0.332497\tvalidation_1-mlogloss:0.783472\n",
      "[71]\tvalidation_0-merror:0.332651\tvalidation_0-mlogloss:0.782854\tvalidation_1-merror:0.332497\tvalidation_1-mlogloss:0.783417\n",
      "[72]\tvalidation_0-merror:0.332651\tvalidation_0-mlogloss:0.782854\tvalidation_1-merror:0.332497\tvalidation_1-mlogloss:0.783407\n",
      "[73]\tvalidation_0-merror:0.332651\tvalidation_0-mlogloss:0.782753\tvalidation_1-merror:0.332497\tvalidation_1-mlogloss:0.783351\n",
      "[74]\tvalidation_0-merror:0.332485\tvalidation_0-mlogloss:0.782701\tvalidation_1-merror:0.332352\tvalidation_1-mlogloss:0.78333\n",
      "[75]\tvalidation_0-merror:0.332485\tvalidation_0-mlogloss:0.782656\tvalidation_1-merror:0.332352\tvalidation_1-mlogloss:0.78332\n",
      "[76]\tvalidation_0-merror:0.332485\tvalidation_0-mlogloss:0.782595\tvalidation_1-merror:0.332352\tvalidation_1-mlogloss:0.783323\n",
      "[77]\tvalidation_0-merror:0.332485\tvalidation_0-mlogloss:0.782556\tvalidation_1-merror:0.332352\tvalidation_1-mlogloss:0.783246\n",
      "[78]\tvalidation_0-merror:0.332485\tvalidation_0-mlogloss:0.782579\tvalidation_1-merror:0.332352\tvalidation_1-mlogloss:0.783232\n",
      "[79]\tvalidation_0-merror:0.332485\tvalidation_0-mlogloss:0.782462\tvalidation_1-merror:0.332352\tvalidation_1-mlogloss:0.78326\n",
      "[80]\tvalidation_0-merror:0.332485\tvalidation_0-mlogloss:0.78248\tvalidation_1-merror:0.332352\tvalidation_1-mlogloss:0.783178\n",
      "[81]\tvalidation_0-merror:0.332485\tvalidation_0-mlogloss:0.782422\tvalidation_1-merror:0.332352\tvalidation_1-mlogloss:0.783128\n",
      "[82]\tvalidation_0-merror:0.332485\tvalidation_0-mlogloss:0.782419\tvalidation_1-merror:0.332352\tvalidation_1-mlogloss:0.783075\n",
      "[83]\tvalidation_0-merror:0.332485\tvalidation_0-mlogloss:0.782423\tvalidation_1-merror:0.332352\tvalidation_1-mlogloss:0.783116\n",
      "[84]\tvalidation_0-merror:0.332485\tvalidation_0-mlogloss:0.782414\tvalidation_1-merror:0.332352\tvalidation_1-mlogloss:0.783056\n",
      "[85]\tvalidation_0-merror:0.332485\tvalidation_0-mlogloss:0.782387\tvalidation_1-merror:0.332352\tvalidation_1-mlogloss:0.782996\n",
      "[86]\tvalidation_0-merror:0.332485\tvalidation_0-mlogloss:0.782381\tvalidation_1-merror:0.332352\tvalidation_1-mlogloss:0.782957\n",
      "[87]\tvalidation_0-merror:0.332485\tvalidation_0-mlogloss:0.782389\tvalidation_1-merror:0.332352\tvalidation_1-mlogloss:0.782912\n",
      "[88]\tvalidation_0-merror:0.332485\tvalidation_0-mlogloss:0.782242\tvalidation_1-merror:0.332352\tvalidation_1-mlogloss:0.782917\n",
      "[89]\tvalidation_0-merror:0.332527\tvalidation_0-mlogloss:0.782201\tvalidation_1-merror:0.332159\tvalidation_1-mlogloss:0.782876\n",
      "[90]\tvalidation_0-merror:0.332527\tvalidation_0-mlogloss:0.782191\tvalidation_1-merror:0.332159\tvalidation_1-mlogloss:0.782847\n",
      "[91]\tvalidation_0-merror:0.332527\tvalidation_0-mlogloss:0.782174\tvalidation_1-merror:0.332159\tvalidation_1-mlogloss:0.782842\n",
      "[92]\tvalidation_0-merror:0.332527\tvalidation_0-mlogloss:0.782099\tvalidation_1-merror:0.332159\tvalidation_1-mlogloss:0.782792\n",
      "[93]\tvalidation_0-merror:0.332527\tvalidation_0-mlogloss:0.782114\tvalidation_1-merror:0.332159\tvalidation_1-mlogloss:0.782775\n",
      "[94]\tvalidation_0-merror:0.332072\tvalidation_0-mlogloss:0.782126\tvalidation_1-merror:0.332449\tvalidation_1-mlogloss:0.782795\n",
      "[95]\tvalidation_0-merror:0.332072\tvalidation_0-mlogloss:0.782136\tvalidation_1-merror:0.332449\tvalidation_1-mlogloss:0.782787\n",
      "[96]\tvalidation_0-merror:0.332527\tvalidation_0-mlogloss:0.782098\tvalidation_1-merror:0.332159\tvalidation_1-mlogloss:0.782717\n",
      "[97]\tvalidation_0-merror:0.332527\tvalidation_0-mlogloss:0.782054\tvalidation_1-merror:0.332159\tvalidation_1-mlogloss:0.782734\n",
      "[98]\tvalidation_0-merror:0.332072\tvalidation_0-mlogloss:0.782096\tvalidation_1-merror:0.332449\tvalidation_1-mlogloss:0.782738\n",
      "[99]\tvalidation_0-merror:0.332072\tvalidation_0-mlogloss:0.782074\tvalidation_1-merror:0.332449\tvalidation_1-mlogloss:0.782711\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "XGBClassifier(objective='multi:softprob')"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = XGBClassifier()\n",
    "eval_set = [(X_train, y_train), (X_test, y_test)]\n",
    "model.fit(X_train, y_train, eval_metric=[\"merror\", \"mlogloss\"], eval_set=eval_set, verbose=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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": [
    "results = model.evals_result()\n",
    "epochs = len(results['validation_0']['merror'])\n",
    "x_axis = range(0, epochs)\n",
    "# plot log loss\n",
    "fig, ax = plt.subplots()\n",
    "ax.plot(x_axis, results['validation_0']['mlogloss'], label='Train')\n",
    "ax.plot(x_axis, results['validation_1']['mlogloss'], label='Test')\n",
    "ax.legend()\n",
    "plt.ylabel('Log Loss')\n",
    "plt.title('XGBoost Log Loss')\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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": [
    "fig, ax = plt.subplots()\n",
    "ax.plot(x_axis, results['validation_0']['merror'], label='Train')\n",
    "ax.plot(x_axis, results['validation_1']['merror'], label='Test')\n",
    "ax.legend()\n",
    "plt.ylabel('Classification Error')\n",
    "plt.title('XGBoost Classification Error')\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}