{ "cells": [ { "cell_type": "markdown", "metadata": { "id": "i3nScBkEfDn5" }, "source": [ "# Install and Load Packages" ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "r7rp9-CtfGio", "outputId": "281a358e-6327-405e-9e7e-22e369134684" }, "outputs": [], "source": [ "# !pip install flaml" ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "hX-mivp1hQC2", "outputId": "87c45e07-f96d-404e-d8f7-75790038bf07" }, "outputs": [], "source": [ "# !pip install lime" ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "id": "mitElpzkL3ss" }, "outputs": [], "source": [ "import lime\n", "import numpy as np\n", "import pandas as pd\n", "import matplotlib.pyplot as plt\n", "# import seaborn as sns\n", "\n", "import plotly.express as px\n", "import plotly.graph_objects as go\n", "from flaml import AutoML" ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "id": "YzRYyylBMR4B" }, "outputs": [], "source": [ "raw_data = pd.read_csv(\"https://raw.githubusercontent.com/hadimaster65555/dataset_for_teaching/main/dataset/hr_analytics_turnover_dataset/HR_comma_sep.csv\")" ] }, { "cell_type": "markdown", "metadata": { "id": "s8rkIZWPN-4N" }, "source": [ "# Data Inspection" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 226 }, "id": "x6flf6A5NQJj", "outputId": "c1458c04-35b1-411c-ae69-a72222b33101" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
satisfaction_levellast_evaluationnumber_projectaverage_montly_hourstime_spend_companyWork_accidentleftpromotion_last_5yearssalessalary
00.380.5321573010saleslow
10.800.8652626010salesmedium
20.110.8872724010salesmedium
30.720.8752235010saleslow
40.370.5221593010saleslow
\n", "
" ], "text/plain": [ " satisfaction_level last_evaluation number_project average_montly_hours \\\n", "0 0.38 0.53 2 157 \n", "1 0.80 0.86 5 262 \n", "2 0.11 0.88 7 272 \n", "3 0.72 0.87 5 223 \n", "4 0.37 0.52 2 159 \n", "\n", " time_spend_company Work_accident left promotion_last_5years sales \\\n", "0 3 0 1 0 sales \n", "1 6 0 1 0 sales \n", "2 4 0 1 0 sales \n", "3 5 0 1 0 sales \n", "4 3 0 1 0 sales \n", "\n", " salary \n", "0 low \n", "1 medium \n", "2 medium \n", "3 low \n", "4 low " ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "raw_data.head()" ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "q7DEIE8WN8Am", "outputId": "daf3c088-e235-48bd-9143-f81ab47e0103" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "RangeIndex: 14999 entries, 0 to 14998\n", "Data columns (total 10 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 satisfaction_level 14999 non-null float64\n", " 1 last_evaluation 14999 non-null float64\n", " 2 number_project 14999 non-null int64 \n", " 3 average_montly_hours 14999 non-null int64 \n", " 4 time_spend_company 14999 non-null int64 \n", " 5 Work_accident 14999 non-null int64 \n", " 6 left 14999 non-null int64 \n", " 7 promotion_last_5years 14999 non-null int64 \n", " 8 sales 14999 non-null object \n", " 9 salary 14999 non-null object \n", "dtypes: float64(2), int64(6), object(2)\n", "memory usage: 1.1+ MB\n" ] } ], "source": [ "raw_data.info()" ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "2uXPh7iPOFtv", "outputId": "a82c312f-7b90-40a2-f0ea-16896674d28c" }, "outputs": [ { "data": { "text/plain": [ "satisfaction_level 0\n", "last_evaluation 0\n", "number_project 0\n", "average_montly_hours 0\n", "time_spend_company 0\n", "Work_accident 0\n", "left 0\n", "promotion_last_5years 0\n", "sales 0\n", "salary 0\n", "dtype: int64" ] }, "execution_count": 8, "metadata": {}, "output_type": "execute_result" } ], "source": [ "raw_data.isna().sum()" ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "0A6XLrjbOJDI", "outputId": "db8e0a52-dad0-4419-cc8f-091fbe4c1133" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "satisfaction_level: [0.38 0.8 0.11 0.72 0.37 0.41 0.1 0.92 0.89 0.42 0.45 0.84 0.36 0.78\n", " 0.76 0.09 0.46 0.4 0.82 0.87 0.57 0.43 0.13 0.44 0.39 0.85 0.81 0.9\n", " 0.74 0.79 0.17 0.24 0.91 0.71 0.86 0.14 0.75 0.7 0.31 0.73 0.83 0.32\n", " 0.54 0.27 0.77 0.88 0.48 0.19 0.6 0.12 0.61 0.33 0.56 0.47 0.28 0.55\n", " 0.53 0.59 0.66 0.25 0.34 0.58 0.51 0.35 0.64 0.5 0.23 0.15 0.49 0.3\n", " 0.63 0.21 0.62 0.29 0.2 0.16 0.65 0.68 0.67 0.22 0.26 0.99 0.98 1.\n", " 0.52 0.93 0.97 0.69 0.94 0.96 0.18 0.95]\n", "last_evaluation: [0.53 0.86 0.88 0.87 0.52 0.5 0.77 0.85 1. 0.54 0.81 0.92 0.55 0.56\n", " 0.47 0.99 0.51 0.89 0.83 0.95 0.57 0.49 0.46 0.62 0.94 0.48 0.8 0.74\n", " 0.7 0.78 0.91 0.93 0.98 0.97 0.79 0.59 0.84 0.45 0.96 0.68 0.82 0.9\n", " 0.71 0.6 0.65 0.58 0.72 0.67 0.75 0.73 0.63 0.61 0.76 0.66 0.69 0.37\n", " 0.64 0.39 0.41 0.43 0.44 0.36 0.38 0.4 0.42]\n", "number_project: [2 5 7 6 4 3]\n", "average_montly_hours: [157 262 272 223 159 153 247 259 224 142 135 305 234 148 137 143 160 255\n", " 282 147 304 139 158 242 239 128 132 294 134 145 140 246 126 306 152 269\n", " 127 281 276 182 273 307 309 225 226 308 244 286 161 264 277 275 149 295\n", " 151 249 291 232 130 129 155 265 279 284 221 154 150 267 257 177 144 289\n", " 258 263 251 133 216 300 138 260 183 250 292 283 245 256 278 240 136 301\n", " 243 296 274 164 146 261 285 141 297 156 287 219 254 228 131 252 236 270\n", " 298 192 248 266 238 229 233 268 231 253 302 271 290 235 293 241 218 199\n", " 180 195 237 227 172 206 181 217 310 214 198 211 222 213 202 184 204 288\n", " 220 299 303 212 196 179 205 230 203 280 169 188 178 175 166 163 168 165\n", " 189 162 215 193 176 191 174 201 208 171 111 104 106 100 194 209 185 200\n", " 207 187 210 186 167 108 122 110 115 197 102 109 190 99 101 97 173 121\n", " 170 105 118 119 117 114 96 98 107 123 116 125 113 120 112 124 103]\n", "time_spend_company: [ 3 6 4 5 2 8 10 7]\n", "Work_accident: [0 1]\n", "left: [1 0]\n", "promotion_last_5years: [0 1]\n", "sales: ['sales' 'accounting' 'hr' 'technical' 'support' 'management' 'IT'\n", " 'product_mng' 'marketing' 'RandD']\n", "salary: ['low' 'medium' 'high']\n" ] } ], "source": [ "for column in raw_data.columns:\n", " print(f\"{column}:\", raw_data[column].unique())" ] }, { "cell_type": "markdown", "metadata": { "id": "ErgEZpPNOhM7" }, "source": [ "# Data Preprocessing" ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "id": "xVGSRkGPOzKK" }, "outputs": [], "source": [ "raw_data.rename(columns={\"sales\": \"roles\"}, inplace=True)" ] }, { "cell_type": "code", "execution_count": 11, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "fvKTZmGccQ13", "outputId": "f6536755-1e66-4c0b-c12a-85017525fcf8" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "RangeIndex: 14999 entries, 0 to 14998\n", "Data columns (total 10 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 satisfaction_level 14999 non-null float64\n", " 1 last_evaluation 14999 non-null float64\n", " 2 number_project 14999 non-null int64 \n", " 3 average_montly_hours 14999 non-null int64 \n", " 4 time_spend_company 14999 non-null int64 \n", " 5 Work_accident 14999 non-null int64 \n", " 6 left 14999 non-null int64 \n", " 7 promotion_last_5years 14999 non-null int64 \n", " 8 roles 14999 non-null object \n", " 9 salary 14999 non-null object \n", "dtypes: float64(2), int64(6), object(2)\n", "memory usage: 1.1+ MB\n" ] } ], "source": [ "raw_data.info()" ] }, { "cell_type": "code", "execution_count": 12, "metadata": { "id": "U_okl0AhcAi3" }, "outputs": [], "source": [ "X_data = raw_data.drop('left', axis=1).copy()\n", "y_data = raw_data['left'].copy()" ] }, { "cell_type": "code", "execution_count": 13, "metadata": { "id": "qE2KjjhkmCmE" }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "/var/folders/8m/3p682pp13l36tzbh7_bgx2100000gn/T/ipykernel_80949/2785502729.py:1: FutureWarning: Downcasting behavior in `replace` is deprecated and will be removed in a future version. To retain the old behavior, explicitly call `result.infer_objects(copy=False)`. To opt-in to the future behavior, set `pd.set_option('future.no_silent_downcasting', True)`\n", " X_data['salary'] = X_data['salary'].replace({\"low\": 1, \"medium\": 2, \"high\": 3})\n" ] } ], "source": [ "X_data['salary'] = X_data['salary'].replace({\"low\": 1, \"medium\": 2, \"high\": 3})" ] }, { "cell_type": "code", "execution_count": 14, "metadata": { "id": "XrXh9f-9mVoz" }, "outputs": [], "source": [ "X_data = pd.get_dummies(X_data, dtype=float).drop(columns=\"roles_IT\")" ] }, { "cell_type": "code", "execution_count": 15, "metadata": { "id": "hGL1gvFLcqv8" }, "outputs": [], "source": [ "from sklearn.model_selection import train_test_split, GridSearchCV\n", "from sklearn.metrics import classification_report, confusion_matrix" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "Jez9P2A-qxef" }, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": 16, "metadata": { "id": "mzc7I5Z2gTkJ" }, "outputs": [], "source": [ "X_train, X_test, y_train, y_test = train_test_split(\n", " X_data,\n", " y_data,\n", " test_size=0.3,\n", " random_state=65555,\n", " stratify=y_data\n", ")" ] }, { "cell_type": "markdown", "metadata": { "id": "x0WNYCNpROL5" }, "source": [ "# Data Exploration" ] }, { "cell_type": "code", "execution_count": 17, "metadata": { "id": "F7NlEmHsRrvG" }, "outputs": [], "source": [ "# X_data['left'] = y_train.copy()" ] }, { "cell_type": "code", "execution_count": 18, "metadata": { "id": "t4hExItVR1MQ" }, "outputs": [], "source": [ "# X_data['left'].value_counts()" ] }, { "cell_type": "markdown", "metadata": { "id": "jow61cpSdhWy" }, "source": [ "# Modeling" ] }, { "cell_type": "markdown", "metadata": { "id": "PLKzk0mLdSut" }, "source": [ "## Define Model" ] }, { "cell_type": "code", "execution_count": 19, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "qVofOm9ydlA3", "outputId": "30a21528-f865-4785-b666-78d1ec5c2f65" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[flaml.automl.logger: 04-05 01:35:43] {1680} INFO - task = classification\n", "[flaml.automl.logger: 04-05 01:35:43] {1691} INFO - Evaluation method: holdout\n", "[flaml.automl.logger: 04-05 01:35:43] {1789} INFO - Minimizing error metric: 1-accuracy\n", "[flaml.automl.logger: 04-05 01:35:43] {1901} INFO - List of ML learners in AutoML Run: ['xgboost']\n", "[flaml.automl.logger: 04-05 01:35:43] {2219} INFO - iteration 0, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:43] {2345} INFO - Estimated sufficient time budget=293s. Estimated necessary time budget=0s.\n", "[flaml.automl.logger: 04-05 01:35:43] {2392} INFO - at 0.1s,\testimator xgboost's best error=0.2386,\tbest estimator xgboost's best error=0.2386\n", "[flaml.automl.logger: 04-05 01:35:43] {2219} INFO - iteration 1, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:43] {2392} INFO - at 0.1s,\testimator xgboost's best error=0.2386,\tbest estimator xgboost's best error=0.2386\n", "[flaml.automl.logger: 04-05 01:35:43] {2219} INFO - iteration 2, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:43] {2392} INFO - at 0.2s,\testimator xgboost's best error=0.0228,\tbest estimator xgboost's best error=0.0228\n", "[flaml.automl.logger: 04-05 01:35:43] {2219} INFO - iteration 3, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:43] {2392} INFO - at 0.3s,\testimator xgboost's best error=0.0171,\tbest estimator xgboost's best error=0.0171\n", "[flaml.automl.logger: 04-05 01:35:43] {2219} INFO - iteration 4, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:43] {2392} INFO - at 0.3s,\testimator xgboost's best error=0.0171,\tbest estimator xgboost's best error=0.0171\n", "[flaml.automl.logger: 04-05 01:35:43] {2219} INFO - iteration 5, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:43] {2392} INFO - at 0.3s,\testimator xgboost's best error=0.0162,\tbest estimator xgboost's best error=0.0162\n", "[flaml.automl.logger: 04-05 01:35:43] {2219} INFO - iteration 6, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:43] {2392} INFO - at 0.4s,\testimator xgboost's best error=0.0162,\tbest estimator xgboost's best error=0.0162\n", "[flaml.automl.logger: 04-05 01:35:43] {2219} INFO - iteration 7, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:43] {2392} INFO - at 0.5s,\testimator xgboost's best error=0.0162,\tbest estimator xgboost's best error=0.0162\n", "[flaml.automl.logger: 04-05 01:35:43] {2219} INFO - iteration 8, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:43] {2392} INFO - at 0.6s,\testimator xgboost's best error=0.0162,\tbest estimator xgboost's best error=0.0162\n", "[flaml.automl.logger: 04-05 01:35:43] {2219} INFO - iteration 9, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:43] {2392} INFO - at 0.7s,\testimator xgboost's best error=0.0162,\tbest estimator xgboost's best error=0.0162\n", "[flaml.automl.logger: 04-05 01:35:43] {2219} INFO - iteration 10, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:43] {2392} INFO - at 0.8s,\testimator xgboost's best error=0.0162,\tbest estimator xgboost's best error=0.0162\n", "[flaml.automl.logger: 04-05 01:35:43] {2219} INFO - iteration 11, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:43] {2392} INFO - at 0.9s,\testimator xgboost's best error=0.0162,\tbest estimator xgboost's best error=0.0162\n", "[flaml.automl.logger: 04-05 01:35:43] {2219} INFO - iteration 12, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:44] {2392} INFO - at 1.0s,\testimator xgboost's best error=0.0133,\tbest estimator xgboost's best error=0.0133\n", "[flaml.automl.logger: 04-05 01:35:44] {2219} INFO - iteration 13, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:44] {2392} INFO - at 1.1s,\testimator xgboost's best error=0.0133,\tbest estimator xgboost's best error=0.0133\n", "[flaml.automl.logger: 04-05 01:35:44] {2219} INFO - iteration 14, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:44] {2392} INFO - at 1.3s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:44] {2219} INFO - iteration 15, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:44] {2392} INFO - at 1.6s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:44] {2219} INFO - iteration 16, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:44] {2392} INFO - at 1.7s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:44] {2219} INFO - iteration 17, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:44] {2392} INFO - at 1.9s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:44] {2219} INFO - iteration 18, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:45] {2392} INFO - at 2.0s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:45] {2219} INFO - iteration 19, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:45] {2392} INFO - at 2.6s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:45] {2219} INFO - iteration 20, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:45] {2392} INFO - at 2.6s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:45] {2219} INFO - iteration 21, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:45] {2392} INFO - at 2.8s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:45] {2219} INFO - iteration 22, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:45] {2392} INFO - at 3.0s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:45] {2219} INFO - iteration 23, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:46] {2392} INFO - at 3.2s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:46] {2219} INFO - iteration 24, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:46] {2392} INFO - at 3.3s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:46] {2219} INFO - iteration 25, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:46] {2392} INFO - at 3.6s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:46] {2219} INFO - iteration 26, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:46] {2392} INFO - at 3.7s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:46] {2219} INFO - iteration 27, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:46] {2392} INFO - at 3.8s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:46] {2219} INFO - iteration 28, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:47] {2392} INFO - at 4.2s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:47] {2219} INFO - iteration 29, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:47] {2392} INFO - at 4.3s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:47] {2219} INFO - iteration 30, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:47] {2392} INFO - at 4.5s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:47] {2219} INFO - iteration 31, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:47] {2392} INFO - at 4.6s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:47] {2219} INFO - iteration 32, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:47] {2392} INFO - at 4.9s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:47] {2219} INFO - iteration 33, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:48] {2392} INFO - at 5.1s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:48] {2219} INFO - iteration 34, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:48] {2392} INFO - at 5.3s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:48] {2219} INFO - iteration 35, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:48] {2392} INFO - at 5.3s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:48] {2219} INFO - iteration 36, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:49] {2392} INFO - at 6.0s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:49] {2219} INFO - iteration 37, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:49] {2392} INFO - at 6.2s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:49] {2219} INFO - iteration 38, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:49] {2392} INFO - at 6.4s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:49] {2219} INFO - iteration 39, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:49] {2392} INFO - at 6.5s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:49] {2219} INFO - iteration 40, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:49] {2392} INFO - at 6.7s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:49] {2219} INFO - iteration 41, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:49] {2392} INFO - at 6.8s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:49] {2219} INFO - iteration 42, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:50] {2392} INFO - at 7.2s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:50] {2219} INFO - iteration 43, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:50] {2392} INFO - at 7.4s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:50] {2219} INFO - iteration 44, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:50] {2392} INFO - at 7.6s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:50] {2219} INFO - iteration 45, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:50] {2392} INFO - at 7.7s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:50] {2219} INFO - iteration 46, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:50] {2392} INFO - at 7.9s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:50] {2219} INFO - iteration 47, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:51] {2392} INFO - at 8.0s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:51] {2219} INFO - iteration 48, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:51] {2392} INFO - at 8.3s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:51] {2219} INFO - iteration 49, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:51] {2392} INFO - at 8.6s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:51] {2219} INFO - iteration 50, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:51] {2392} INFO - at 8.7s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:51] {2219} INFO - iteration 51, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:51] {2392} INFO - at 8.8s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:51] {2219} INFO - iteration 52, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:52] {2392} INFO - at 9.3s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:52] {2219} INFO - iteration 53, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:52] {2392} INFO - at 9.4s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:52] {2219} INFO - iteration 54, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:52] {2392} INFO - at 9.6s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:52] {2219} INFO - iteration 55, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:52] {2392} INFO - at 9.7s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:52] {2219} INFO - iteration 56, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:53] {2392} INFO - at 10.1s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:53] {2219} INFO - iteration 57, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:53] {2392} INFO - at 10.7s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:53] {2219} INFO - iteration 58, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:53] {2392} INFO - at 10.7s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:53] {2219} INFO - iteration 59, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:54] {2392} INFO - at 11.2s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:54] {2219} INFO - iteration 60, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:54] {2392} INFO - at 11.2s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:54] {2219} INFO - iteration 61, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:54] {2392} INFO - at 11.9s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:54] {2219} INFO - iteration 62, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:54] {2392} INFO - at 12.0s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:54] {2219} INFO - iteration 63, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:55] {2392} INFO - at 12.0s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:55] {2219} INFO - iteration 64, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:55] {2392} INFO - at 12.6s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:55] {2219} INFO - iteration 65, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:55] {2392} INFO - at 12.6s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:55] {2219} INFO - iteration 66, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:56] {2392} INFO - at 13.2s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:56] {2219} INFO - iteration 67, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:56] {2392} INFO - at 13.8s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:56] {2219} INFO - iteration 68, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:56] {2392} INFO - at 13.8s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:56] {2219} INFO - iteration 69, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:57] {2392} INFO - at 14.5s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:57] {2219} INFO - iteration 70, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:57] {2392} INFO - at 14.5s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:57] {2219} INFO - iteration 71, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:58] {2392} INFO - at 15.1s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:58] {2219} INFO - iteration 72, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:58] {2392} INFO - at 15.2s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:58] {2219} INFO - iteration 73, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:58] {2392} INFO - at 15.4s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:58] {2219} INFO - iteration 74, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:58] {2392} INFO - at 15.5s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:58] {2219} INFO - iteration 75, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:58] {2392} INFO - at 15.8s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:58] {2219} INFO - iteration 76, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:58] {2392} INFO - at 15.9s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:58] {2219} INFO - iteration 77, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:59] {2392} INFO - at 16.9s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:59] {2219} INFO - iteration 78, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:35:59] {2392} INFO - at 17.0s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:35:59] {2219} INFO - iteration 79, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:00] {2392} INFO - at 17.4s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:36:00] {2219} INFO - iteration 80, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:00] {2392} INFO - at 17.5s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:36:00] {2219} INFO - iteration 81, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:01] {2392} INFO - at 18.1s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:36:01] {2219} INFO - iteration 82, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:01] {2392} INFO - at 18.1s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:36:01] {2219} INFO - iteration 83, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:01] {2392} INFO - at 18.4s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:36:01] {2219} INFO - iteration 84, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:01] {2392} INFO - at 18.5s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:36:01] {2219} INFO - iteration 85, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:01] {2392} INFO - at 18.6s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:36:01] {2219} INFO - iteration 86, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:01] {2392} INFO - at 18.9s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:36:01] {2219} INFO - iteration 87, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:02] {2392} INFO - at 19.2s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:36:02] {2219} INFO - iteration 88, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:02] {2392} INFO - at 19.3s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:36:02] {2219} INFO - iteration 89, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:02] {2392} INFO - at 19.5s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:36:02] {2219} INFO - iteration 90, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:02] {2392} INFO - at 19.7s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:36:02] {2219} INFO - iteration 91, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:02] {2392} INFO - at 19.7s,\testimator xgboost's best error=0.0086,\tbest estimator xgboost's best error=0.0086\n", "[flaml.automl.logger: 04-05 01:36:02] {2219} INFO - iteration 92, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:03] {2392} INFO - at 20.3s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:03] {2219} INFO - iteration 93, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:03] {2392} INFO - at 20.5s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:03] {2219} INFO - iteration 94, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:04] {2392} INFO - at 21.0s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:04] {2219} INFO - iteration 95, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:04] {2392} INFO - at 21.4s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:04] {2219} INFO - iteration 96, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:04] {2392} INFO - at 21.8s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:04] {2219} INFO - iteration 97, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:05] {2392} INFO - at 22.2s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:05] {2219} INFO - iteration 98, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:05] {2392} INFO - at 22.9s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:05] {2219} INFO - iteration 99, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:07] {2392} INFO - at 24.0s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:07] {2219} INFO - iteration 100, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:07] {2392} INFO - at 24.2s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:07] {2219} INFO - iteration 101, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:08] {2392} INFO - at 25.1s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:08] {2219} INFO - iteration 102, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:08] {2392} INFO - at 25.4s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:08] {2219} INFO - iteration 103, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:08] {2392} INFO - at 25.6s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:08] {2219} INFO - iteration 104, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:09] {2392} INFO - at 26.0s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:09] {2219} INFO - iteration 105, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:10] {2392} INFO - at 27.6s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:10] {2219} INFO - iteration 106, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:10] {2392} INFO - at 27.8s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:10] {2219} INFO - iteration 107, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:11] {2392} INFO - at 28.2s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:11] {2219} INFO - iteration 108, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:11] {2392} INFO - at 28.9s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:11] {2219} INFO - iteration 109, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:12] {2392} INFO - at 29.4s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:12] {2219} INFO - iteration 110, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:12] {2392} INFO - at 29.9s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:12] {2219} INFO - iteration 111, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:13] {2392} INFO - at 30.1s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:13] {2219} INFO - iteration 112, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:13] {2392} INFO - at 30.5s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:13] {2219} INFO - iteration 113, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:14] {2392} INFO - at 31.9s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:14] {2219} INFO - iteration 114, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:15] {2392} INFO - at 32.0s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:15] {2219} INFO - iteration 115, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:15] {2392} INFO - at 32.4s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:15] {2219} INFO - iteration 116, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:15] {2392} INFO - at 32.6s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:15] {2219} INFO - iteration 117, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:16] {2392} INFO - at 33.4s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:16] {2219} INFO - iteration 118, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:16] {2392} INFO - at 33.6s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:16] {2219} INFO - iteration 119, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:17] {2392} INFO - at 34.4s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:17] {2219} INFO - iteration 120, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:17] {2392} INFO - at 34.8s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:17] {2219} INFO - iteration 121, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:18] {2392} INFO - at 35.0s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:18] {2219} INFO - iteration 122, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:19] {2392} INFO - at 36.0s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:19] {2219} INFO - iteration 123, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:19] {2392} INFO - at 36.4s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:19] {2219} INFO - iteration 124, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:19] {2392} INFO - at 36.9s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:19] {2219} INFO - iteration 125, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:20] {2392} INFO - at 37.2s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:20] {2219} INFO - iteration 126, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:20] {2392} INFO - at 37.7s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:20] {2219} INFO - iteration 127, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:21] {2392} INFO - at 38.4s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:21] {2219} INFO - iteration 128, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:21] {2392} INFO - at 38.8s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:21] {2219} INFO - iteration 129, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:22] {2392} INFO - at 39.0s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:22] {2219} INFO - iteration 130, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:22] {2392} INFO - at 39.5s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:22] {2219} INFO - iteration 131, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:22] {2392} INFO - at 39.8s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:22] {2219} INFO - iteration 132, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:23] {2392} INFO - at 40.3s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:23] {2219} INFO - iteration 133, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:23] {2392} INFO - at 40.6s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:23] {2219} INFO - iteration 134, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:24] {2392} INFO - at 41.4s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:24] {2219} INFO - iteration 135, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:25] {2392} INFO - at 42.2s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:25] {2219} INFO - iteration 136, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:25] {2392} INFO - at 42.5s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:25] {2219} INFO - iteration 137, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:26] {2392} INFO - at 43.0s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:26] {2219} INFO - iteration 138, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:26] {2392} INFO - at 43.3s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:26] {2219} INFO - iteration 139, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:27] {2392} INFO - at 44.0s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:27] {2219} INFO - iteration 140, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:27] {2392} INFO - at 44.4s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:27] {2219} INFO - iteration 141, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:27] {2392} INFO - at 44.9s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:27] {2219} INFO - iteration 142, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:28] {2392} INFO - at 45.5s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:28] {2219} INFO - iteration 143, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:28] {2392} INFO - at 45.9s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:28] {2219} INFO - iteration 144, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:30] {2392} INFO - at 47.4s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:30] {2219} INFO - iteration 145, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:30] {2392} INFO - at 47.7s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:30] {2219} INFO - iteration 146, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:31] {2392} INFO - at 48.4s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:31] {2219} INFO - iteration 147, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:32] {2392} INFO - at 49.1s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:32] {2219} INFO - iteration 148, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:32] {2392} INFO - at 49.2s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:32] {2219} INFO - iteration 149, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:32] {2392} INFO - at 49.4s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:32] {2219} INFO - iteration 150, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:32] {2392} INFO - at 49.9s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:32] {2219} INFO - iteration 151, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:33] {2392} INFO - at 50.3s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:33] {2219} INFO - iteration 152, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:33] {2392} INFO - at 50.7s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:33] {2219} INFO - iteration 153, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:34] {2392} INFO - at 51.1s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:34] {2219} INFO - iteration 154, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:34] {2392} INFO - at 51.5s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:34] {2219} INFO - iteration 155, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:35] {2392} INFO - at 52.1s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:35] {2219} INFO - iteration 156, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:35] {2392} INFO - at 52.3s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:35] {2219} INFO - iteration 157, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:35] {2392} INFO - at 52.7s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:35] {2219} INFO - iteration 158, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:36] {2392} INFO - at 53.0s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:36] {2219} INFO - iteration 159, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:36] {2392} INFO - at 53.2s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:36] {2219} INFO - iteration 160, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:37] {2392} INFO - at 54.4s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:37] {2219} INFO - iteration 161, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:37] {2392} INFO - at 54.7s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:37] {2219} INFO - iteration 162, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:38] {2392} INFO - at 55.1s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:38] {2219} INFO - iteration 163, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:38] {2392} INFO - at 55.3s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:38] {2219} INFO - iteration 164, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:39] {2392} INFO - at 56.8s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:39] {2219} INFO - iteration 165, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:40] {2392} INFO - at 57.4s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:40] {2219} INFO - iteration 166, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:40] {2392} INFO - at 57.6s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:40] {2219} INFO - iteration 167, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:41] {2392} INFO - at 58.3s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:41] {2219} INFO - iteration 168, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:41] {2392} INFO - at 58.6s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:41] {2219} INFO - iteration 169, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:41] {2392} INFO - at 58.8s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:41] {2219} INFO - iteration 170, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:42] {2392} INFO - at 59.2s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:42] {2219} INFO - iteration 171, current learner xgboost\n", "[flaml.automl.logger: 04-05 01:36:42] {2392} INFO - at 59.8s,\testimator xgboost's best error=0.0076,\tbest estimator xgboost's best error=0.0076\n", "[flaml.automl.logger: 04-05 01:36:43] {2628} INFO - retrain xgboost for 0.5s\n", "[flaml.automl.logger: 04-05 01:36:43] {2631} INFO - retrained model: XGBClassifier(base_score=None, booster=None, callbacks=[],\n", " colsample_bylevel=0.5210664820067598, colsample_bynode=None,\n", " colsample_bytree=0.8326041578837515, device=None,\n", " early_stopping_rounds=None, enable_categorical=False,\n", " eval_metric=None, feature_types=None, gamma=None,\n", " grow_policy='lossguide', importance_type=None,\n", " interaction_constraints=None, learning_rate=0.6064978559327864,\n", " max_bin=None, max_cat_threshold=None, max_cat_to_onehot=None,\n", " max_delta_step=None, max_depth=0, max_leaves=211,\n", " min_child_weight=0.0888416394086547, missing=nan,\n", " monotone_constraints=None, multi_strategy=None, n_estimators=18,\n", " n_jobs=-1, num_parallel_tree=None, random_state=None, ...)\n", "[flaml.automl.logger: 04-05 01:36:43] {1931} INFO - fit succeeded\n", "[flaml.automl.logger: 04-05 01:36:43] {1932} INFO - Time taken to find the best model: 20.277902126312256\n" ] } ], "source": [ "automl = AutoML()\n", "settings = {\n", " \"time_budget\": 60, # total running time in seconds\n", " \"metric\": \"accuracy\", # primary metrics for regression can be chosen from: ['mae','mse','r2']\n", " \"estimator_list\": [\n", " \"xgboost\"\n", " ], # list of ML learners; we tune XGBoost in this example\n", " \"task\": \"classification\", # task type\n", " \"log_file_name\": \"houses_experiment.log\", # flaml log file\n", " \"seed\": 7654321, # random seed\n", "}\n", "automl.fit(X_train=X_train, y_train=y_train, **settings)\n", "\n", "# automl.fit(X_train, y_train, task=\"classification\", )" ] }, { "cell_type": "code", "execution_count": 20, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 35 }, "id": "ToDR5D7qdCqq", "outputId": "c0c82beb-718c-4544-9ddb-a429561174ae" }, "outputs": [ { "data": { "text/plain": [ "'xgboost'" ] }, "execution_count": 20, "metadata": {}, "output_type": "execute_result" } ], "source": [ "automl.best_estimator" ] }, { "cell_type": "code", "execution_count": 21, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 252 }, "id": "nP8pkjL0d4AF", "outputId": "1b1117d3-914a-4426-9260-c437278bfe46" }, "outputs": [ { "data": { "text/html": [ "
XGBClassifier(base_score=None, booster=None, callbacks=[],\n",
       "              colsample_bylevel=0.5210664820067598, colsample_bynode=None,\n",
       "              colsample_bytree=0.8326041578837515, device=None,\n",
       "              early_stopping_rounds=None, enable_categorical=False,\n",
       "              eval_metric=None, feature_types=None, gamma=None,\n",
       "              grow_policy='lossguide', importance_type=None,\n",
       "              interaction_constraints=None, learning_rate=0.6064978559327864,\n",
       "              max_bin=None, max_cat_threshold=None, max_cat_to_onehot=None,\n",
       "              max_delta_step=None, max_depth=0, max_leaves=211,\n",
       "              min_child_weight=0.0888416394086547, missing=nan,\n",
       "              monotone_constraints=None, multi_strategy=None, n_estimators=18,\n",
       "              n_jobs=-1, num_parallel_tree=None, random_state=None, ...)
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
" ], "text/plain": [ "XGBClassifier(base_score=None, booster=None, callbacks=[],\n", " colsample_bylevel=0.5210664820067598, colsample_bynode=None,\n", " colsample_bytree=0.8326041578837515, device=None,\n", " early_stopping_rounds=None, enable_categorical=False,\n", " eval_metric=None, feature_types=None, gamma=None,\n", " grow_policy='lossguide', importance_type=None,\n", " interaction_constraints=None, learning_rate=0.6064978559327864,\n", " max_bin=None, max_cat_threshold=None, max_cat_to_onehot=None,\n", " max_delta_step=None, max_depth=0, max_leaves=211,\n", " min_child_weight=0.0888416394086547, missing=nan,\n", " monotone_constraints=None, multi_strategy=None, n_estimators=18,\n", " n_jobs=-1, num_parallel_tree=None, random_state=None, ...)" ] }, "execution_count": 21, "metadata": {}, "output_type": "execute_result" } ], "source": [ "automl.model.estimator" ] }, { "cell_type": "code", "execution_count": 22, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 448 }, "id": "X6CdSuibeI-q", "outputId": "119a9939-542b-477d-d30a-257a94071c31" }, "outputs": [ { "data": { "text/plain": [ "" ] }, "execution_count": 22, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "import matplotlib.pyplot as plt\n", "\n", "plt.barh(automl.feature_names_in_, automl.feature_importances_)" ] }, { "cell_type": "code", "execution_count": 23, "metadata": { "id": "AGrRO2p1ecfp" }, "outputs": [], "source": [ "y_pred = automl.predict_proba(X_train)" ] }, { "cell_type": "code", "execution_count": 24, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "kdIuOXYYv2y8", "outputId": "8d0b7110-b85f-4874-f712-d854074a0594" }, "outputs": [ { "data": { "text/plain": [ "array([[9.9942148e-01, 5.7852472e-04],\n", " [2.1207333e-03, 9.9787927e-01],\n", " [9.9984801e-01, 1.5201201e-04],\n", " ...,\n", " [9.7779787e-01, 2.2202114e-02],\n", " [9.9904227e-01, 9.5770403e-04],\n", " [9.9985534e-01, 1.4465630e-04]], dtype=float32)" ] }, "execution_count": 24, "metadata": {}, "output_type": "execute_result" } ], "source": [ "y_pred" ] }, { "cell_type": "code", "execution_count": 25, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "1C2UmBrIf2I5", "outputId": "bc39c5e8-c45b-4916-9234-de8bb73d9239" }, "outputs": [ { "data": { "text/plain": [ "array([[9.9942148e-01, 5.7852472e-04],\n", " [2.1207333e-03, 9.9787927e-01],\n", " [9.9984801e-01, 1.5201201e-04],\n", " ...,\n", " [9.7779787e-01, 2.2202114e-02],\n", " [9.9904227e-01, 9.5770403e-04],\n", " [9.9985534e-01, 1.4465630e-04]], dtype=float32)" ] }, "execution_count": 25, "metadata": {}, "output_type": "execute_result" } ], "source": [ "y_pred" ] }, { "cell_type": "code", "execution_count": 26, "metadata": { "id": "vYzK2kHKeilJ" }, "outputs": [], "source": [ "from flaml.automl.ml import sklearn_metric_loss_score" ] }, { "cell_type": "code", "execution_count": 27, "metadata": { "id": "Bn7ejeK3iOtw" }, "outputs": [], "source": [ "from lime import lime_tabular" ] }, { "cell_type": "code", "execution_count": 28, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "IpISelHFpChe", "outputId": "03b8d6e6-3c51-42ed-8acf-87405d52e04b" }, "outputs": [ { "data": { "text/plain": [ "satisfaction_level 0.80\n", "last_evaluation 0.86\n", "number_project 5.00\n", "average_montly_hours 262.00\n", "time_spend_company 6.00\n", "Work_accident 0.00\n", "promotion_last_5years 0.00\n", "salary 2.00\n", "roles_RandD 0.00\n", "roles_accounting 0.00\n", "roles_hr 0.00\n", "roles_management 0.00\n", "roles_marketing 0.00\n", "roles_product_mng 0.00\n", "roles_sales 1.00\n", "roles_support 0.00\n", "roles_technical 0.00\n", "Name: 1, dtype: float64" ] }, "execution_count": 28, "metadata": {}, "output_type": "execute_result" } ], "source": [ "X_train.loc[1,:]" ] }, { "cell_type": "code", "execution_count": 29, "metadata": { "id": "D6NBFF6MklBf" }, "outputs": [], "source": [ "from lime.lime_tabular import LimeTabularExplainer\n", "explainer = LimeTabularExplainer(X_train,\n", " feature_names=X_train.columns,\n", " class_names=['stay', 'left'],\n", " discretize_continuous=False,\n", " verbose=True)" ] }, { "cell_type": "code", "execution_count": 30, "metadata": { "id": "rhzVjemGoQqH" }, "outputs": [], "source": [ "user_id_1 = X_train.iloc[1]" ] }, { "cell_type": "code", "execution_count": 31, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "AHvhuH4pkod8", "outputId": "48eab42a-4881-42ca-b2d3-e723d7657acc" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Intercept 0.03966242624928315\n", "Prediction_local [0.03388112]\n", "Right: 0.99787927\n" ] } ], "source": [ "lime_res = explainer.explain_instance(user_id_1, automl.predict_proba, num_features=len(X_train.columns))" ] }, { "cell_type": "code", "execution_count": 32, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 452 }, "id": "nQK270UfAj3C", "outputId": "cc3c684c-499b-4d7b-80b1-8765aad68ae2" }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "lime_res.as_pyplot_figure();" ] }, { "cell_type": "code", "execution_count": 33, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 534 }, "id": "QeD73ale_XYZ", "outputId": "52e5151b-3618-469c-80f7-7bd6cd4c0dd2" }, "outputs": [ { "data": { "text/html": [ "\n", " \n", " \n", "
\n", " \n", " \n", " " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "lime_res.show_in_notebook(show_table=True, show_all=True)" ] }, { "cell_type": "code", "execution_count": 34, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "nPeT5b306rHi", "outputId": "c8b7d5fd-41ef-4d58-c84e-90136eef2e8d" }, "outputs": [ { "data": { "text/plain": [ "[('satisfaction_level', -0.047379411147259035),\n", " ('average_montly_hours', 0.03552324913767401),\n", " ('last_evaluation', 0.006114148508735551),\n", " ('time_spend_company', 0.004848226327314694),\n", " ('roles_sales', 0.0037604610042036835),\n", " ('roles_hr', -0.003378530603008509),\n", " ('salary', -0.003161843918446249),\n", " ('promotion_last_5years', -0.0030093554155614815),\n", " ('roles_management', -0.0025003673235702586),\n", " ('roles_product_mng', 0.002185821544424093),\n", " ('roles_accounting', -0.0016521773684327504),\n", " ('roles_technical', -0.0016168585772014452),\n", " ('number_project', 0.0015916946196413149),\n", " ('roles_support', -0.00153374095025741),\n", " ('roles_marketing', 0.0010454562878820077),\n", " ('roles_RandD', 0.0009804297656538794),\n", " ('Work_accident', -0.000721277811929021)]" ] }, "execution_count": 34, "metadata": {}, "output_type": "execute_result" } ], "source": [ "lime_res.as_list()" ] }, { "cell_type": "code", "execution_count": 35, "metadata": { "id": "Y4pnC2Q_68_h" }, "outputs": [], "source": [ "iseng = pd.DataFrame(lime_res.as_list()).rename(columns={0: \"variable\", 1: \"value\"})" ] }, { "cell_type": "code", "execution_count": 36, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 542 }, "id": "VwLxcvsJAJi6", "outputId": "caaeb01a-96e2-46b7-adee-c79ce1237607" }, "outputs": [ { "data": { "text/html": [ " \n", " " ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "application/vnd.plotly.v1+json": { "config": { "plotlyServerURL": "https://plot.ly" }, "data": [ { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "satisfaction_level", "marker": { "color": "#636efa", "pattern": { "shape": "" } }, "name": "satisfaction_level", "offsetgroup": "satisfaction_level", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ -0.047379411147259035 ], "xaxis": "x", "y": [ "satisfaction_level" ], "yaxis": "y" }, { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "average_montly_hours", "marker": { "color": "#EF553B", "pattern": { "shape": "" } }, "name": "average_montly_hours", "offsetgroup": "average_montly_hours", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ 0.03552324913767401 ], "xaxis": "x", "y": [ "average_montly_hours" ], "yaxis": "y" }, { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "last_evaluation", "marker": { "color": "#00cc96", "pattern": { "shape": "" } }, "name": "last_evaluation", "offsetgroup": "last_evaluation", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ 0.006114148508735551 ], "xaxis": "x", "y": [ "last_evaluation" ], "yaxis": "y" }, { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "time_spend_company", "marker": { "color": "#ab63fa", "pattern": { "shape": "" } }, "name": "time_spend_company", "offsetgroup": "time_spend_company", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ 0.004848226327314694 ], "xaxis": "x", "y": [ "time_spend_company" ], "yaxis": "y" }, { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "roles_sales", "marker": { "color": "#FFA15A", "pattern": { "shape": "" } }, "name": "roles_sales", "offsetgroup": "roles_sales", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ 0.0037604610042036835 ], "xaxis": "x", "y": [ "roles_sales" ], "yaxis": "y" }, { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "roles_hr", "marker": { "color": "#19d3f3", "pattern": { "shape": "" } }, "name": "roles_hr", "offsetgroup": "roles_hr", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ -0.003378530603008509 ], "xaxis": "x", "y": [ "roles_hr" ], "yaxis": "y" }, { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "salary", "marker": { "color": "#FF6692", "pattern": { "shape": "" } }, "name": "salary", "offsetgroup": "salary", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ -0.003161843918446249 ], "xaxis": "x", "y": [ "salary" ], "yaxis": "y" }, { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "promotion_last_5years", "marker": { "color": "#B6E880", "pattern": { "shape": "" } }, "name": "promotion_last_5years", "offsetgroup": "promotion_last_5years", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ -0.0030093554155614815 ], "xaxis": "x", "y": [ "promotion_last_5years" ], "yaxis": "y" }, { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "roles_management", "marker": { "color": "#FF97FF", "pattern": { "shape": "" } }, "name": "roles_management", "offsetgroup": "roles_management", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ -0.0025003673235702586 ], "xaxis": "x", "y": [ "roles_management" ], "yaxis": "y" }, { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "roles_product_mng", "marker": { "color": "#FECB52", "pattern": { "shape": "" } }, "name": "roles_product_mng", "offsetgroup": "roles_product_mng", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ 0.002185821544424093 ], "xaxis": "x", "y": [ "roles_product_mng" ], "yaxis": "y" }, { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "roles_accounting", "marker": { "color": "#636efa", "pattern": { "shape": "" } }, "name": "roles_accounting", "offsetgroup": "roles_accounting", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ -0.0016521773684327504 ], "xaxis": "x", "y": [ "roles_accounting" ], "yaxis": "y" }, { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "roles_technical", "marker": { "color": "#EF553B", "pattern": { "shape": "" } }, "name": "roles_technical", "offsetgroup": "roles_technical", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ -0.0016168585772014452 ], "xaxis": "x", "y": [ "roles_technical" ], "yaxis": "y" }, { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "number_project", "marker": { "color": "#00cc96", "pattern": { "shape": "" } }, "name": "number_project", "offsetgroup": "number_project", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ 0.0015916946196413149 ], "xaxis": "x", "y": [ "number_project" ], "yaxis": "y" }, { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "roles_support", "marker": { "color": "#ab63fa", "pattern": { "shape": "" } }, "name": "roles_support", "offsetgroup": "roles_support", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ -0.00153374095025741 ], "xaxis": "x", "y": [ "roles_support" ], "yaxis": "y" }, { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "roles_marketing", "marker": { "color": "#FFA15A", "pattern": { "shape": "" } }, "name": "roles_marketing", "offsetgroup": "roles_marketing", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ 0.0010454562878820077 ], "xaxis": "x", "y": [ "roles_marketing" ], "yaxis": "y" }, { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "roles_RandD", "marker": { "color": "#19d3f3", "pattern": { "shape": "" } }, "name": "roles_RandD", "offsetgroup": "roles_RandD", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ 0.0009804297656538794 ], "xaxis": "x", "y": [ "roles_RandD" ], "yaxis": "y" }, { "alignmentgroup": "True", "hovertemplate": "variable=%{y}
value=%{x}", "legendgroup": "Work_accident", "marker": { "color": "#FF6692", "pattern": { "shape": "" } }, "name": "Work_accident", "offsetgroup": "Work_accident", "orientation": "h", "showlegend": true, "textposition": "auto", "type": "bar", "x": [ -0.000721277811929021 ], "xaxis": "x", "y": [ "Work_accident" ], "yaxis": "y" } ], "layout": { "autosize": true, "barmode": "relative", "legend": { "title": { "text": "variable" }, "tracegroupgap": 0 }, "margin": { "t": 60 }, "template": { "data": { "bar": [ { "error_x": { "color": "#2a3f5f" }, "error_y": { "color": "#2a3f5f" }, "marker": { "line": { "color": "#E5ECF6", "width": 0.5 }, "pattern": { "fillmode": "overlay", "size": 10, "solidity": 0.2 } }, "type": "bar" } ], "barpolar": [ { "marker": { "line": { "color": "#E5ECF6", "width": 0.5 }, "pattern": { "fillmode": "overlay", "size": 10, "solidity": 0.2 } }, "type": "barpolar" } ], "carpet": [ { "aaxis": { "endlinecolor": "#2a3f5f", "gridcolor": "white", "linecolor": "white", "minorgridcolor": "white", "startlinecolor": "#2a3f5f" }, "baxis": { "endlinecolor": "#2a3f5f", "gridcolor": "white", "linecolor": "white", "minorgridcolor": "white", "startlinecolor": "#2a3f5f" }, "type": "carpet" } ], "choropleth": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "type": "choropleth" } ], "contour": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1, "#f0f921" ] ], "type": "contour" } ], "contourcarpet": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "type": "contourcarpet" } ], "heatmap": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1, "#f0f921" ] ], "type": "heatmap" } ], "heatmapgl": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1, "#f0f921" ] ], "type": "heatmapgl" } ], "histogram": [ { "marker": { "pattern": { "fillmode": "overlay", "size": 10, "solidity": 0.2 } }, "type": "histogram" } ], "histogram2d": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1, "#f0f921" ] ], "type": "histogram2d" } ], "histogram2dcontour": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1, "#f0f921" ] ], "type": "histogram2dcontour" } ], "mesh3d": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "type": "mesh3d" } ], "parcoords": [ { "line": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "parcoords" } ], "pie": [ { "automargin": true, "type": "pie" } ], "scatter": [ { "fillpattern": { "fillmode": "overlay", "size": 10, "solidity": 0.2 }, "type": "scatter" } ], "scatter3d": [ { "line": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scatter3d" } ], "scattercarpet": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scattercarpet" } ], "scattergeo": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scattergeo" } ], "scattergl": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scattergl" } ], "scattermapbox": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scattermapbox" } ], "scatterpolar": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scatterpolar" } ], "scatterpolargl": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scatterpolargl" } ], "scatterternary": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scatterternary" } ], "surface": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1, "#f0f921" ] ], "type": "surface" } ], "table": [ { "cells": { "fill": { "color": "#EBF0F8" }, "line": { "color": "white" } }, "header": { "fill": { "color": "#C8D4E3" }, "line": { "color": "white" } }, "type": "table" } ] }, "layout": { "annotationdefaults": { "arrowcolor": "#2a3f5f", "arrowhead": 0, "arrowwidth": 1 }, "autotypenumbers": "strict", "coloraxis": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "colorscale": { "diverging": [ [ 0, "#8e0152" ], [ 0.1, "#c51b7d" ], [ 0.2, "#de77ae" ], [ 0.3, "#f1b6da" ], [ 0.4, "#fde0ef" ], [ 0.5, "#f7f7f7" ], [ 0.6, "#e6f5d0" ], [ 0.7, "#b8e186" ], [ 0.8, "#7fbc41" ], [ 0.9, "#4d9221" ], [ 1, "#276419" ] ], "sequential": [ [ 0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1, "#f0f921" ] ], "sequentialminus": [ [ 0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1, "#f0f921" ] ] }, "colorway": [ "#636efa", "#EF553B", "#00cc96", "#ab63fa", "#FFA15A", "#19d3f3", "#FF6692", "#B6E880", "#FF97FF", "#FECB52" ], "font": { "color": "#2a3f5f" }, "geo": { "bgcolor": "white", "lakecolor": "white", "landcolor": "#E5ECF6", "showlakes": true, "showland": true, "subunitcolor": "white" }, "hoverlabel": { "align": "left" }, "hovermode": "closest", "mapbox": { "style": "light" }, "paper_bgcolor": "white", "plot_bgcolor": "#E5ECF6", "polar": { "angularaxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" }, "bgcolor": "#E5ECF6", "radialaxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" } }, "scene": { "xaxis": { "backgroundcolor": "#E5ECF6", "gridcolor": "white", "gridwidth": 2, "linecolor": "white", "showbackground": true, "ticks": "", "zerolinecolor": "white" }, "yaxis": { "backgroundcolor": "#E5ECF6", "gridcolor": "white", "gridwidth": 2, "linecolor": "white", "showbackground": true, "ticks": "", "zerolinecolor": "white" }, "zaxis": { "backgroundcolor": "#E5ECF6", "gridcolor": "white", "gridwidth": 2, "linecolor": "white", "showbackground": true, "ticks": "", "zerolinecolor": "white" } }, "shapedefaults": { "line": { "color": "#2a3f5f" } }, "ternary": { "aaxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" }, "baxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" }, "bgcolor": "#E5ECF6", "caxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" } }, "title": { "x": 0.05 }, "xaxis": { "automargin": true, "gridcolor": "white", "linecolor": "white", "ticks": "", "title": { "standoff": 15 }, "zerolinecolor": "white", "zerolinewidth": 2 }, "yaxis": { "automargin": true, "gridcolor": "white", "linecolor": "white", "ticks": "", "title": { "standoff": 15 }, "zerolinecolor": "white", "zerolinewidth": 2 } } }, "xaxis": { "anchor": "y", "autorange": true, "domain": [ 0, 1 ], "range": [ -0.051985114496421986, 0.040128952486836955 ], "title": { "text": "value" }, "type": "linear" }, "yaxis": { "anchor": "x", "autorange": true, "categoryarray": [ "Work_accident", "roles_RandD", "roles_marketing", "roles_support", "number_project", "roles_technical", "roles_accounting", "roles_product_mng", "roles_management", "promotion_last_5years", "salary", "roles_hr", "roles_sales", "time_spend_company", "last_evaluation", "average_montly_hours", "satisfaction_level" ], "categoryorder": "array", "domain": [ 0, 1 ], "range": [ -0.5, 16.5 ], "title": { "text": "variable" }, "type": "category" } } }, "image/png": "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", "text/html": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "fig = px.bar(\n", " iseng,\n", " y=\"variable\",\n", " x=\"value\",\n", " color=\"variable\", orientation=\"h\")\n", "\n", "fig.show()" ] }, { "cell_type": "code", "execution_count": 37, "metadata": { "id": "WpelG3ZW8K7j" }, "outputs": [], "source": [ "import pickle\n", "with open('automl.pkl', 'wb') as f:\n", " pickle.dump(automl, f, pickle.HIGHEST_PROTOCOL)" ] }, { "cell_type": "code", "execution_count": 82, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "M_QtsE13HSDW", "outputId": "25417059-6019-48b5-eae0-1c7818acfe0d" }, "outputs": [ { "data": { "text/plain": [ "" ] }, "execution_count": 82, "metadata": {}, "output_type": "execute_result" } ], "source": [ "explainer" ] }, { "cell_type": "code", "execution_count": 38, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "nUvCK1NN1agY", "outputId": "9ed7d854-fd9f-4ae8-f442-eac08b5f3f8b" }, "outputs": [ { "data": { "text/plain": [ "array([[0.00726569, 0.9927343 ]], dtype=float32)" ] }, "execution_count": 38, "metadata": {}, "output_type": "execute_result" } ], "source": [ "automl.predict_proba(\n", " pd.DataFrame(\n", " [user_id_1.values.tolist()],\n", " columns=user_id_1.index.tolist()\n", " )\n", ")" ] }, { "cell_type": "code", "execution_count": 39, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "7w2hFUJz2K6g", "outputId": "1a857c41-1b5d-4103-a198-726145c3b472" }, "outputs": [ { "data": { "text/plain": [ "1" ] }, "execution_count": 39, "metadata": {}, "output_type": "execute_result" } ], "source": [ "y_train.iloc[1]" ] }, { "cell_type": "code", "execution_count": 40, "metadata": { "id": "nzBAxeu6tDDx" }, "outputs": [], "source": [ "colname_list = X_train.columns.to_list()\n", "container = []\n", "for i in range(len(colname_list)):\n", " container.append((i, colname_list[i]))\n", "\n", "container = dict(container)\n", "\n", "explainer_table = pd.DataFrame(lime_res.local_exp[1]).replace(container).rename(columns={0: \"variable\", 1: \"value\"}).copy()" ] }, { "cell_type": "code", "execution_count": 41, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 582 }, "id": "uMKjJ8CauUkL", "outputId": "fa4dc694-66eb-4d67-d7bd-e95ad8320a71" }, "outputs": [ { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "summary": "{\n \"name\": \"explainer_table\",\n \"rows\": 17,\n \"fields\": [\n {\n \"column\": \"variable\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 17,\n \"samples\": [\n \"satisfaction_level\",\n \"average_montly_hours\",\n \"roles_RandD\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"value\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 0.01368503840260982,\n \"min\": -0.043170091696445916,\n \"max\": 0.02788512115936838,\n \"num_unique_values\": 17,\n \"samples\": [\n -0.043170091696445916,\n 0.02788512115936838,\n -0.004311957227516555\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}", "type": "dataframe", "variable_name": "explainer_table" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
variablevalue
0satisfaction_level-0.043170
1average_montly_hours0.027885
2time_spend_company0.014083
3last_evaluation0.008947
4number_project-0.004481
5roles_RandD-0.004312
6roles_hr-0.004029
7roles_sales0.002918
8roles_management0.001900
9salary-0.001476
10Work_accident0.001409
11roles_accounting-0.001405
12roles_marketing-0.001164
13roles_support0.001077
14promotion_last_5years-0.000965
15roles_product_mng-0.000964
16roles_technical-0.000688
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "
\n", "
\n" ], "text/plain": [ " variable value\n", "0 satisfaction_level -0.043170\n", "1 average_montly_hours 0.027885\n", "2 time_spend_company 0.014083\n", "3 last_evaluation 0.008947\n", "4 number_project -0.004481\n", "5 roles_RandD -0.004312\n", "6 roles_hr -0.004029\n", "7 roles_sales 0.002918\n", "8 roles_management 0.001900\n", "9 salary -0.001476\n", "10 Work_accident 0.001409\n", "11 roles_accounting -0.001405\n", "12 roles_marketing -0.001164\n", "13 roles_support 0.001077\n", "14 promotion_last_5years -0.000965\n", "15 roles_product_mng -0.000964\n", "16 roles_technical -0.000688" ] }, "execution_count": 41, "metadata": {}, "output_type": "execute_result" } ], "source": [ "explainer_table" ] }, { "cell_type": "code", "execution_count": 68, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 365 }, "id": "Kk_h1rM5uVtI", "outputId": "19f343c4-c324-41d4-ebef-56da2d615e6e" }, "outputs": [ { "ename": "ValueError", "evalue": "Value of 'x' is not the name of a column in 'data_frame'. Expected one of [0, 1] but received: value", "output_type": "error", "traceback": [ "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", "\u001b[0;31mValueError\u001b[0m Traceback (most recent call last)", "\u001b[0;32m\u001b[0m in \u001b[0;36m\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m fig = px.bar(\n\u001b[0m\u001b[1;32m 2\u001b[0m \u001b[0mlime_res\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mas_list\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[1;32m 3\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"variable\"\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 4\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"value\"\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 5\u001b[0m color=\"variable\", orientation=\"h\")\n", "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/plotly/express/_chart_types.py\u001b[0m in \u001b[0;36mbar\u001b[0;34m(data_frame, x, y, color, pattern_shape, facet_row, facet_col, facet_col_wrap, facet_row_spacing, facet_col_spacing, hover_name, hover_data, custom_data, text, base, error_x, error_x_minus, error_y, error_y_minus, animation_frame, animation_group, category_orders, labels, color_discrete_sequence, color_discrete_map, color_continuous_scale, pattern_shape_sequence, pattern_shape_map, range_color, color_continuous_midpoint, opacity, orientation, barmode, log_x, log_y, range_x, range_y, text_auto, title, template, width, height)\u001b[0m\n\u001b[1;32m 371\u001b[0m \u001b[0mmark\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 372\u001b[0m \"\"\"\n\u001b[0;32m--> 373\u001b[0;31m return make_figure(\n\u001b[0m\u001b[1;32m 374\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mlocals\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[1;32m 375\u001b[0m \u001b[0mconstructor\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mgo\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mBar\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/plotly/express/_core.py\u001b[0m in \u001b[0;36mmake_figure\u001b[0;34m(args, constructor, trace_patch, layout_patch)\u001b[0m\n\u001b[1;32m 2001\u001b[0m \u001b[0mapply_default_cascade\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 2002\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2003\u001b[0;31m \u001b[0margs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mbuild_dataframe\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mconstructor\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 2004\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mconstructor\u001b[0m \u001b[0;32min\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mgo\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mTreemap\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mgo\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mSunburst\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mgo\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mIcicle\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"path\"\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 2005\u001b[0m \u001b[0margs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mprocess_dataframe_hierarchy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/plotly/express/_core.py\u001b[0m in \u001b[0;36mbuild_dataframe\u001b[0;34m(args, constructor)\u001b[0m\n\u001b[1;32m 1410\u001b[0m \u001b[0;31m# now that things have been prepped, we do the systematic rewriting of `args`\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1411\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1412\u001b[0;31m df_output, wide_id_vars = process_args_into_dataframe(\n\u001b[0m\u001b[1;32m 1413\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mwide_mode\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvar_name\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvalue_name\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1414\u001b[0m )\n", "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/plotly/express/_core.py\u001b[0m in \u001b[0;36mprocess_args_into_dataframe\u001b[0;34m(args, wide_mode, var_name, value_name)\u001b[0m\n\u001b[1;32m 1206\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0margument\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;34m\"index\"\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1207\u001b[0m \u001b[0merr_msg\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0;34m\"\\n To use the index, pass it in directly as `df.index`.\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1208\u001b[0;31m \u001b[0;32mraise\u001b[0m \u001b[0mValueError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0merr_msg\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 1209\u001b[0m \u001b[0;32melif\u001b[0m \u001b[0mlength\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdf_input\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0margument\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m!=\u001b[0m \u001b[0mlength\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 1210\u001b[0m raise ValueError(\n", "\u001b[0;31mValueError\u001b[0m: Value of 'x' is not the name of a column in 'data_frame'. Expected one of [0, 1] but received: value" ] } ], "source": [ "fig = px.bar(\n", " lime_res.as_list(),\n", " y=\"variable\",\n", " x=\"value\",\n", " color=\"variable\", orientation=\"h\")\n", "\n", "fig.show()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "C3d9TJKYuwUx" }, "outputs": [], "source": [ "user_id_2 = X_train.iloc[2]" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "kGPRp-VrvWdJ" }, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "5ZYd9iWNxKyN" }, "outputs": [], "source": [ "pred_result = automl.predict_proba(X_train.iloc[[2]])[0]\n", "pred_result" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "lXmvXCmt2YMA" }, "outputs": [], "source": [ "def raw_pred_to_class_pred(data):\n", " employee_att_stat = [\"stay\", \"left\"]\n", " pred_result = automl.predict_proba(data)[0]\n", " return dict(zip(employee_att_stat, pred_result))" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "efOJ_yGB3sCv" }, "outputs": [], "source": [ "raw_pred_to_class_pred(X_train.iloc[[2]])" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "Dk-iEDzH3zJO" }, "outputs": [], "source": [ "def lime_visualizer(data, model):\n", " lime_res = explainer.explain_instance(data, model.predict_proba)\n", " explainer_table = pd.DataFrame(lime_res.as_list()).rename(columns={0: \"variable\", 1: \"value\"})\n", " fig = px.bar(\n", " explainer_table,\n", " y=\"variable\",\n", " x=\"value\",\n", " color=\"variable\", orientation=\"h\")\n", " return fig" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "7-Ch_AsVwVQL" }, "outputs": [], "source": [ "result = lime_visualizer(X_train.iloc[2], automl)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "CS9zPuXewPNB" }, "outputs": [], "source": [ "result" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "Mc1oSqTcwo-J" }, "outputs": [], "source": [] } ], "metadata": { "colab": { "provenance": [] }, "kernelspec": { "display_name": "Python 3 (ipykernel)", "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": "3.11.8" } }, "nbformat": 4, "nbformat_minor": 4 }