├── .gitattributes
├── .ipynb_checkpoints
    ├── bank churn model - modeling(GBDT)-checkpoint.ipynb
    ├── bank churn model - preprocessing-checkpoint.ipynb
    └── bank churn model - preview-checkpoint.ipynb
├── ExternalData.csv
├── bank churn model - modeling(GBDT).ipynb
├── bank churn model - preprocessing.ipynb
├── bank churn model - preview.ipynb
├── bankChurn.csv
└── model_data.csv


/.gitattributes:
--------------------------------------------------------------------------------
1 | # Auto detect text files and perform LF normalization
2 | * text=auto
3 | 


--------------------------------------------------------------------------------
/.ipynb_checkpoints/bank churn model - modeling(GBDT)-checkpoint.ipynb:
--------------------------------------------------------------------------------
  1 | {
  2 |  "cells": [
  3 |   {
  4 |    "cell_type": "code",
  5 |    "execution_count": 1,
  6 |    "metadata": {},
  7 |    "outputs": [
  8 |     {
  9 |      "name": "stderr",
 10 |      "output_type": "stream",
 11 |      "text": [
 12 |       "D:\\Anaconda3\\lib\\site-packages\\sklearn\\cross_validation.py:41: DeprecationWarning: This module was deprecated in version 0.18 in favor of the model_selection module into which all the refactored classes and functions are moved. Also note that the interface of the new CV iterators are different from that of this module. This module will be removed in 0.20.\n",
 13 |       "  \"This module will be removed in 0.20.\", DeprecationWarning)\n"
 14 |      ]
 15 |     }
 16 |    ],
 17 |    "source": [
 18 |     "import pandas as pd\n",
 19 |     "from sklearn import ensemble, cross_validation, metrics\n",
 20 |     "from sklearn.ensemble import GradientBoostingClassifier\n",
 21 |     "from sklearn.cross_validation import train_test_split\n",
 22 |     "from sklearn.cross_validation import KFold\n",
 23 |     "from sklearn.model_selection import GridSearchCV"
 24 |    ]
 25 |   },
 26 |   {
 27 |    "cell_type": "code",
 28 |    "execution_count": 2,
 29 |    "metadata": {},
 30 |    "outputs": [
 31 |     {
 32 |      "data": {
 33 |       "text/html": [
 34 |        "<div>\n",
 35 |        "<style>\n",
 36 |        "    .dataframe thead tr:only-child th {\n",
 37 |        "        text-align: right;\n",
 38 |        "    }\n",
 39 |        "\n",
 40 |        "    .dataframe thead th {\n",
 41 |        "        text-align: left;\n",
 42 |        "    }\n",
 43 |        "\n",
 44 |        "    .dataframe tbody tr th {\n",
 45 |        "        vertical-align: top;\n",
 46 |        "    }\n",
 47 |        "</style>\n",
 48 |        "<table border=\"1\" class=\"dataframe\">\n",
 49 |        "  <thead>\n",
 50 |        "    <tr style=\"text-align: right;\">\n",
 51 |        "      <th></th>\n",
 52 |        "      <th>LOCAL_CUR_SAV_SLOPE</th>\n",
 53 |        "      <th>LOCAL_CUR_MON_AVG_BAL</th>\n",
 54 |        "      <th>LOCAL_OVEONEYR_FF_MON_AVG_BAL</th>\n",
 55 |        "      <th>LOCAL_FIX_MON_AVG_BAL</th>\n",
 56 |        "      <th>LOCAL_FIX_MON_AVG_BAL_PROP</th>\n",
 57 |        "      <th>LOCAL_BELONEYR_FF_SLOPE</th>\n",
 58 |        "      <th>LOCAL_BELONEYR_FF_MON_AVG_BAL</th>\n",
 59 |        "      <th>LOCAL_OVEONEYR_FF_SLOPE</th>\n",
 60 |        "      <th>LOCAL_SAV_SLOPE</th>\n",
 61 |        "      <th>LOCAL_SAV_CUR_ALL_BAL</th>\n",
 62 |        "      <th>...</th>\n",
 63 |        "      <th>kid6_10_nan</th>\n",
 64 |        "      <th>kid11_15_U</th>\n",
 65 |        "      <th>kid11_15_Y</th>\n",
 66 |        "      <th>kid11_15_nan</th>\n",
 67 |        "      <th>kid16_17_U</th>\n",
 68 |        "      <th>kid16_17_Y</th>\n",
 69 |        "      <th>kid16_17_nan</th>\n",
 70 |        "      <th>car_buy_New</th>\n",
 71 |        "      <th>car_buy_UNKNOWN</th>\n",
 72 |        "      <th>car_buy_nan</th>\n",
 73 |        "    </tr>\n",
 74 |        "  </thead>\n",
 75 |        "  <tbody>\n",
 76 |        "    <tr>\n",
 77 |        "      <th>0</th>\n",
 78 |        "      <td>0.333333</td>\n",
 79 |        "      <td>0.000000</td>\n",
 80 |        "      <td>0.009068</td>\n",
 81 |        "      <td>0.001157</td>\n",
 82 |        "      <td>1.000000</td>\n",
 83 |        "      <td>0.344437</td>\n",
 84 |        "      <td>0.0</td>\n",
 85 |        "      <td>0.333284</td>\n",
 86 |        "      <td>0.333196</td>\n",
 87 |        "      <td>9.277653e-04</td>\n",
 88 |        "      <td>...</td>\n",
 89 |        "      <td>0.0</td>\n",
 90 |        "      <td>1.0</td>\n",
 91 |        "      <td>0.0</td>\n",
 92 |        "      <td>0.0</td>\n",
 93 |        "      <td>1.0</td>\n",
 94 |        "      <td>0.0</td>\n",
 95 |        "      <td>0.0</td>\n",
 96 |        "      <td>1.0</td>\n",
 97 |        "      <td>0.0</td>\n",
 98 |        "      <td>0.0</td>\n",
 99 |        "    </tr>\n",
100 |        "    <tr>\n",
101 |        "      <th>1</th>\n",
102 |        "      <td>0.668564</td>\n",
103 |        "      <td>0.002648</td>\n",
104 |        "      <td>0.010078</td>\n",
105 |        "      <td>0.001286</td>\n",
106 |        "      <td>0.735565</td>\n",
107 |        "      <td>0.344437</td>\n",
108 |        "      <td>0.0</td>\n",
109 |        "      <td>0.293169</td>\n",
110 |        "      <td>0.345716</td>\n",
111 |        "      <td>7.225051e-04</td>\n",
112 |        "      <td>...</td>\n",
113 |        "      <td>0.0</td>\n",
114 |        "      <td>1.0</td>\n",
115 |        "      <td>0.0</td>\n",
116 |        "      <td>0.0</td>\n",
117 |        "      <td>1.0</td>\n",
118 |        "      <td>0.0</td>\n",
119 |        "      <td>0.0</td>\n",
120 |        "      <td>1.0</td>\n",
121 |        "      <td>0.0</td>\n",
122 |        "      <td>0.0</td>\n",
123 |        "    </tr>\n",
124 |        "    <tr>\n",
125 |        "      <th>2</th>\n",
126 |        "      <td>0.333521</td>\n",
127 |        "      <td>0.000002</td>\n",
128 |        "      <td>0.000000</td>\n",
129 |        "      <td>0.000000</td>\n",
130 |        "      <td>0.000000</td>\n",
131 |        "      <td>0.344437</td>\n",
132 |        "      <td>0.0</td>\n",
133 |        "      <td>0.344442</td>\n",
134 |        "      <td>0.344545</td>\n",
135 |        "      <td>4.032385e-07</td>\n",
136 |        "      <td>...</td>\n",
137 |        "      <td>0.0</td>\n",
138 |        "      <td>1.0</td>\n",
139 |        "      <td>0.0</td>\n",
140 |        "      <td>0.0</td>\n",
141 |        "      <td>1.0</td>\n",
142 |        "      <td>0.0</td>\n",
143 |        "      <td>0.0</td>\n",
144 |        "      <td>1.0</td>\n",
145 |        "      <td>0.0</td>\n",
146 |        "      <td>0.0</td>\n",
147 |        "    </tr>\n",
148 |        "    <tr>\n",
149 |        "      <th>3</th>\n",
150 |        "      <td>0.149696</td>\n",
151 |        "      <td>0.003858</td>\n",
152 |        "      <td>0.000000</td>\n",
153 |        "      <td>0.000000</td>\n",
154 |        "      <td>0.000000</td>\n",
155 |        "      <td>0.344437</td>\n",
156 |        "      <td>0.0</td>\n",
157 |        "      <td>0.344442</td>\n",
158 |        "      <td>0.154644</td>\n",
159 |        "      <td>7.529284e-04</td>\n",
160 |        "      <td>...</td>\n",
161 |        "      <td>0.0</td>\n",
162 |        "      <td>1.0</td>\n",
163 |        "      <td>0.0</td>\n",
164 |        "      <td>0.0</td>\n",
165 |        "      <td>1.0</td>\n",
166 |        "      <td>0.0</td>\n",
167 |        "      <td>0.0</td>\n",
168 |        "      <td>0.0</td>\n",
169 |        "      <td>1.0</td>\n",
170 |        "      <td>0.0</td>\n",
171 |        "    </tr>\n",
172 |        "    <tr>\n",
173 |        "      <th>4</th>\n",
174 |        "      <td>0.333429</td>\n",
175 |        "      <td>0.000005</td>\n",
176 |        "      <td>0.000000</td>\n",
177 |        "      <td>0.000000</td>\n",
178 |        "      <td>0.000000</td>\n",
179 |        "      <td>0.344437</td>\n",
180 |        "      <td>0.0</td>\n",
181 |        "      <td>0.344442</td>\n",
182 |        "      <td>0.344450</td>\n",
183 |        "      <td>7.852062e-07</td>\n",
184 |        "      <td>...</td>\n",
185 |        "      <td>0.0</td>\n",
186 |        "      <td>1.0</td>\n",
187 |        "      <td>0.0</td>\n",
188 |        "      <td>0.0</td>\n",
189 |        "      <td>0.0</td>\n",
190 |        "      <td>1.0</td>\n",
191 |        "      <td>0.0</td>\n",
192 |        "      <td>1.0</td>\n",
193 |        "      <td>0.0</td>\n",
194 |        "      <td>0.0</td>\n",
195 |        "    </tr>\n",
196 |        "  </tbody>\n",
197 |        "</table>\n",
198 |        "<p>5 rows × 178 columns</p>\n",
199 |        "</div>"
200 |       ],
201 |       "text/plain": [
202 |        "   LOCAL_CUR_SAV_SLOPE  LOCAL_CUR_MON_AVG_BAL  LOCAL_OVEONEYR_FF_MON_AVG_BAL  \\\n",
203 |        "0             0.333333               0.000000                       0.009068   \n",
204 |        "1             0.668564               0.002648                       0.010078   \n",
205 |        "2             0.333521               0.000002                       0.000000   \n",
206 |        "3             0.149696               0.003858                       0.000000   \n",
207 |        "4             0.333429               0.000005                       0.000000   \n",
208 |        "\n",
209 |        "   LOCAL_FIX_MON_AVG_BAL  LOCAL_FIX_MON_AVG_BAL_PROP  LOCAL_BELONEYR_FF_SLOPE  \\\n",
210 |        "0               0.001157                    1.000000                 0.344437   \n",
211 |        "1               0.001286                    0.735565                 0.344437   \n",
212 |        "2               0.000000                    0.000000                 0.344437   \n",
213 |        "3               0.000000                    0.000000                 0.344437   \n",
214 |        "4               0.000000                    0.000000                 0.344437   \n",
215 |        "\n",
216 |        "   LOCAL_BELONEYR_FF_MON_AVG_BAL  LOCAL_OVEONEYR_FF_SLOPE  LOCAL_SAV_SLOPE  \\\n",
217 |        "0                            0.0                 0.333284         0.333196   \n",
218 |        "1                            0.0                 0.293169         0.345716   \n",
219 |        "2                            0.0                 0.344442         0.344545   \n",
220 |        "3                            0.0                 0.344442         0.154644   \n",
221 |        "4                            0.0                 0.344442         0.344450   \n",
222 |        "\n",
223 |        "   LOCAL_SAV_CUR_ALL_BAL     ...       kid6_10_nan  kid11_15_U  kid11_15_Y  \\\n",
224 |        "0           9.277653e-04     ...               0.0         1.0         0.0   \n",
225 |        "1           7.225051e-04     ...               0.0         1.0         0.0   \n",
226 |        "2           4.032385e-07     ...               0.0         1.0         0.0   \n",
227 |        "3           7.529284e-04     ...               0.0         1.0         0.0   \n",
228 |        "4           7.852062e-07     ...               0.0         1.0         0.0   \n",
229 |        "\n",
230 |        "   kid11_15_nan  kid16_17_U  kid16_17_Y  kid16_17_nan  car_buy_New  \\\n",
231 |        "0           0.0         1.0         0.0           0.0          1.0   \n",
232 |        "1           0.0         1.0         0.0           0.0          1.0   \n",
233 |        "2           0.0         1.0         0.0           0.0          1.0   \n",
234 |        "3           0.0         1.0         0.0           0.0          0.0   \n",
235 |        "4           0.0         0.0         1.0           0.0          1.0   \n",
236 |        "\n",
237 |        "   car_buy_UNKNOWN  car_buy_nan  \n",
238 |        "0              0.0          0.0  \n",
239 |        "1              0.0          0.0  \n",
240 |        "2              0.0          0.0  \n",
241 |        "3              1.0          0.0  \n",
242 |        "4              0.0          0.0  \n",
243 |        "\n",
244 |        "[5 rows x 178 columns]"
245 |       ]
246 |      },
247 |      "execution_count": 2,
248 |      "metadata": {},
249 |      "output_type": "execute_result"
250 |     }
251 |    ],
252 |    "source": [
253 |     "model_data = pd.read_csv('./model_data.csv', encoding='utf-8')\n",
254 |     "model_data.head()"
255 |    ]
256 |   },
257 |   {
258 |    "cell_type": "code",
259 |    "execution_count": 3,
260 |    "metadata": {
261 |     "collapsed": true
262 |    },
263 |    "outputs": [],
264 |    "source": [
265 |     "all_features = list(model_data.columns)\n",
266 |     "all_features.remove('CHURN_CUST_IND')"
267 |    ]
268 |   },
269 |   {
270 |    "cell_type": "code",
271 |    "execution_count": 4,
272 |    "metadata": {
273 |     "collapsed": true
274 |    },
275 |    "outputs": [],
276 |    "source": [
277 |     "X_train,  X_test, y_train, y_test = train_test_split(model_data[all_features], model_data['CHURN_CUST_IND'], test_size=0.3, random_state=1)"
278 |    ]
279 |   },
280 |   {
281 |    "cell_type": "code",
282 |    "execution_count": 5,
283 |    "metadata": {},
284 |    "outputs": [
285 |     {
286 |      "name": "stdout",
287 |      "output_type": "stream",
288 |      "text": [
289 |       "y_train_count:  0.0    10851\n",
290 |       "1.0     1217\n",
291 |       "Name: CHURN_CUST_IND, dtype: int64\n",
292 |       "y_test_count:  0.0    4649\n",
293 |       "1.0     524\n",
294 |       "Name: CHURN_CUST_IND, dtype: int64\n"
295 |      ]
296 |     }
297 |    ],
298 |    "source": [
299 |     "print('y_train_count: ', y_train.value_counts())\n",
300 |     "print('y_test_count: ', y_test.value_counts())"
301 |    ]
302 |   },
303 |   {
304 |    "cell_type": "code",
305 |    "execution_count": 6,
306 |    "metadata": {},
307 |    "outputs": [
308 |     {
309 |      "name": "stdout",
310 |      "output_type": "stream",
311 |      "text": [
312 |       "Accuracy : 0.9115\n",
313 |       "AUC Score (Testing): 0.855843\n"
314 |      ]
315 |     }
316 |    ],
317 |    "source": [
318 |     "# 使用GBDT模块的默认参数进行训练\n",
319 |     "gbm0 = GradientBoostingClassifier(random_state=10)\n",
320 |     "gbm0.fit(X_train,y_train)\n",
321 |     "y_pred = gbm0.predict(X_test)\n",
322 |     "y_predprob = gbm0.predict_proba(X_test)[:,1]\n",
323 |     "print(\"Accuracy : %.4g\" % metrics.accuracy_score(y_test, y_pred))\n",
324 |     "print (\"AUC Score (Testing): %f\" % metrics.roc_auc_score(y_test, y_predprob))"
325 |    ]
326 |   },
327 |   {
328 |    "cell_type": "code",
329 |    "execution_count": null,
330 |    "metadata": {
331 |     "collapsed": true
332 |    },
333 |    "outputs": [],
334 |    "source": []
335 |   }
336 |  ],
337 |  "metadata": {
338 |   "kernelspec": {
339 |    "display_name": "Python 3",
340 |    "language": "python",
341 |    "name": "python3"
342 |   },
343 |   "language_info": {
344 |    "codemirror_mode": {
345 |     "name": "ipython",
346 |     "version": 3
347 |    },
348 |    "file_extension": ".py",
349 |    "mimetype": "text/x-python",
350 |    "name": "python",
351 |    "nbconvert_exporter": "python",
352 |    "pygments_lexer": "ipython3",
353 |    "version": "3.6.3"
354 |   }
355 |  },
356 |  "nbformat": 4,
357 |  "nbformat_minor": 2
358 | }
359 | 


--------------------------------------------------------------------------------
/bank churn model - modeling(GBDT).ipynb:
--------------------------------------------------------------------------------
  1 | {
  2 |  "cells": [
  3 |   {
  4 |    "cell_type": "code",
  5 |    "execution_count": 47,
  6 |    "metadata": {},
  7 |    "outputs": [],
  8 |    "source": [
  9 |     "import pandas as pd\n",
 10 |     "from sklearn import ensemble, cross_validation, metrics\n",
 11 |     "from sklearn.ensemble import GradientBoostingClassifier\n",
 12 |     "from sklearn.cross_validation import train_test_split\n",
 13 |     "from sklearn.cross_validation import KFold\n",
 14 |     "from sklearn.model_selection import GridSearchCV\n",
 15 |     "from sklearn.metrics import roc_curve, auc\n",
 16 |     "import matplotlib.pyplot as plt"
 17 |    ]
 18 |   },
 19 |   {
 20 |    "cell_type": "code",
 21 |    "execution_count": 2,
 22 |    "metadata": {},
 23 |    "outputs": [
 24 |     {
 25 |      "data": {
 26 |       "text/html": [
 27 |        "<div>\n",
 28 |        "<style>\n",
 29 |        "    .dataframe thead tr:only-child th {\n",
 30 |        "        text-align: right;\n",
 31 |        "    }\n",
 32 |        "\n",
 33 |        "    .dataframe thead th {\n",
 34 |        "        text-align: left;\n",
 35 |        "    }\n",
 36 |        "\n",
 37 |        "    .dataframe tbody tr th {\n",
 38 |        "        vertical-align: top;\n",
 39 |        "    }\n",
 40 |        "</style>\n",
 41 |        "<table border=\"1\" class=\"dataframe\">\n",
 42 |        "  <thead>\n",
 43 |        "    <tr style=\"text-align: right;\">\n",
 44 |        "      <th></th>\n",
 45 |        "      <th>LOCAL_CUR_SAV_SLOPE</th>\n",
 46 |        "      <th>LOCAL_CUR_MON_AVG_BAL</th>\n",
 47 |        "      <th>LOCAL_OVEONEYR_FF_MON_AVG_BAL</th>\n",
 48 |        "      <th>LOCAL_FIX_MON_AVG_BAL</th>\n",
 49 |        "      <th>LOCAL_FIX_MON_AVG_BAL_PROP</th>\n",
 50 |        "      <th>LOCAL_BELONEYR_FF_SLOPE</th>\n",
 51 |        "      <th>LOCAL_BELONEYR_FF_MON_AVG_BAL</th>\n",
 52 |        "      <th>LOCAL_OVEONEYR_FF_SLOPE</th>\n",
 53 |        "      <th>LOCAL_SAV_SLOPE</th>\n",
 54 |        "      <th>LOCAL_SAV_CUR_ALL_BAL</th>\n",
 55 |        "      <th>...</th>\n",
 56 |        "      <th>kid6_10_nan</th>\n",
 57 |        "      <th>kid11_15_U</th>\n",
 58 |        "      <th>kid11_15_Y</th>\n",
 59 |        "      <th>kid11_15_nan</th>\n",
 60 |        "      <th>kid16_17_U</th>\n",
 61 |        "      <th>kid16_17_Y</th>\n",
 62 |        "      <th>kid16_17_nan</th>\n",
 63 |        "      <th>car_buy_New</th>\n",
 64 |        "      <th>car_buy_UNKNOWN</th>\n",
 65 |        "      <th>car_buy_nan</th>\n",
 66 |        "    </tr>\n",
 67 |        "  </thead>\n",
 68 |        "  <tbody>\n",
 69 |        "    <tr>\n",
 70 |        "      <th>0</th>\n",
 71 |        "      <td>0.333333</td>\n",
 72 |        "      <td>0.000000</td>\n",
 73 |        "      <td>0.009068</td>\n",
 74 |        "      <td>0.001157</td>\n",
 75 |        "      <td>1.000000</td>\n",
 76 |        "      <td>0.344437</td>\n",
 77 |        "      <td>0.0</td>\n",
 78 |        "      <td>0.333284</td>\n",
 79 |        "      <td>0.333196</td>\n",
 80 |        "      <td>9.277653e-04</td>\n",
 81 |        "      <td>...</td>\n",
 82 |        "      <td>0.0</td>\n",
 83 |        "      <td>1.0</td>\n",
 84 |        "      <td>0.0</td>\n",
 85 |        "      <td>0.0</td>\n",
 86 |        "      <td>1.0</td>\n",
 87 |        "      <td>0.0</td>\n",
 88 |        "      <td>0.0</td>\n",
 89 |        "      <td>1.0</td>\n",
 90 |        "      <td>0.0</td>\n",
 91 |        "      <td>0.0</td>\n",
 92 |        "    </tr>\n",
 93 |        "    <tr>\n",
 94 |        "      <th>1</th>\n",
 95 |        "      <td>0.668564</td>\n",
 96 |        "      <td>0.002648</td>\n",
 97 |        "      <td>0.010078</td>\n",
 98 |        "      <td>0.001286</td>\n",
 99 |        "      <td>0.735565</td>\n",
100 |        "      <td>0.344437</td>\n",
101 |        "      <td>0.0</td>\n",
102 |        "      <td>0.293169</td>\n",
103 |        "      <td>0.345716</td>\n",
104 |        "      <td>7.225051e-04</td>\n",
105 |        "      <td>...</td>\n",
106 |        "      <td>0.0</td>\n",
107 |        "      <td>1.0</td>\n",
108 |        "      <td>0.0</td>\n",
109 |        "      <td>0.0</td>\n",
110 |        "      <td>1.0</td>\n",
111 |        "      <td>0.0</td>\n",
112 |        "      <td>0.0</td>\n",
113 |        "      <td>1.0</td>\n",
114 |        "      <td>0.0</td>\n",
115 |        "      <td>0.0</td>\n",
116 |        "    </tr>\n",
117 |        "    <tr>\n",
118 |        "      <th>2</th>\n",
119 |        "      <td>0.333521</td>\n",
120 |        "      <td>0.000002</td>\n",
121 |        "      <td>0.000000</td>\n",
122 |        "      <td>0.000000</td>\n",
123 |        "      <td>0.000000</td>\n",
124 |        "      <td>0.344437</td>\n",
125 |        "      <td>0.0</td>\n",
126 |        "      <td>0.344442</td>\n",
127 |        "      <td>0.344545</td>\n",
128 |        "      <td>4.032385e-07</td>\n",
129 |        "      <td>...</td>\n",
130 |        "      <td>0.0</td>\n",
131 |        "      <td>1.0</td>\n",
132 |        "      <td>0.0</td>\n",
133 |        "      <td>0.0</td>\n",
134 |        "      <td>1.0</td>\n",
135 |        "      <td>0.0</td>\n",
136 |        "      <td>0.0</td>\n",
137 |        "      <td>1.0</td>\n",
138 |        "      <td>0.0</td>\n",
139 |        "      <td>0.0</td>\n",
140 |        "    </tr>\n",
141 |        "    <tr>\n",
142 |        "      <th>3</th>\n",
143 |        "      <td>0.149696</td>\n",
144 |        "      <td>0.003858</td>\n",
145 |        "      <td>0.000000</td>\n",
146 |        "      <td>0.000000</td>\n",
147 |        "      <td>0.000000</td>\n",
148 |        "      <td>0.344437</td>\n",
149 |        "      <td>0.0</td>\n",
150 |        "      <td>0.344442</td>\n",
151 |        "      <td>0.154644</td>\n",
152 |        "      <td>7.529284e-04</td>\n",
153 |        "      <td>...</td>\n",
154 |        "      <td>0.0</td>\n",
155 |        "      <td>1.0</td>\n",
156 |        "      <td>0.0</td>\n",
157 |        "      <td>0.0</td>\n",
158 |        "      <td>1.0</td>\n",
159 |        "      <td>0.0</td>\n",
160 |        "      <td>0.0</td>\n",
161 |        "      <td>0.0</td>\n",
162 |        "      <td>1.0</td>\n",
163 |        "      <td>0.0</td>\n",
164 |        "    </tr>\n",
165 |        "    <tr>\n",
166 |        "      <th>4</th>\n",
167 |        "      <td>0.333429</td>\n",
168 |        "      <td>0.000005</td>\n",
169 |        "      <td>0.000000</td>\n",
170 |        "      <td>0.000000</td>\n",
171 |        "      <td>0.000000</td>\n",
172 |        "      <td>0.344437</td>\n",
173 |        "      <td>0.0</td>\n",
174 |        "      <td>0.344442</td>\n",
175 |        "      <td>0.344450</td>\n",
176 |        "      <td>7.852062e-07</td>\n",
177 |        "      <td>...</td>\n",
178 |        "      <td>0.0</td>\n",
179 |        "      <td>1.0</td>\n",
180 |        "      <td>0.0</td>\n",
181 |        "      <td>0.0</td>\n",
182 |        "      <td>0.0</td>\n",
183 |        "      <td>1.0</td>\n",
184 |        "      <td>0.0</td>\n",
185 |        "      <td>1.0</td>\n",
186 |        "      <td>0.0</td>\n",
187 |        "      <td>0.0</td>\n",
188 |        "    </tr>\n",
189 |        "  </tbody>\n",
190 |        "</table>\n",
191 |        "<p>5 rows × 178 columns</p>\n",
192 |        "</div>"
193 |       ],
194 |       "text/plain": [
195 |        "   LOCAL_CUR_SAV_SLOPE  LOCAL_CUR_MON_AVG_BAL  LOCAL_OVEONEYR_FF_MON_AVG_BAL  \\\n",
196 |        "0             0.333333               0.000000                       0.009068   \n",
197 |        "1             0.668564               0.002648                       0.010078   \n",
198 |        "2             0.333521               0.000002                       0.000000   \n",
199 |        "3             0.149696               0.003858                       0.000000   \n",
200 |        "4             0.333429               0.000005                       0.000000   \n",
201 |        "\n",
202 |        "   LOCAL_FIX_MON_AVG_BAL  LOCAL_FIX_MON_AVG_BAL_PROP  LOCAL_BELONEYR_FF_SLOPE  \\\n",
203 |        "0               0.001157                    1.000000                 0.344437   \n",
204 |        "1               0.001286                    0.735565                 0.344437   \n",
205 |        "2               0.000000                    0.000000                 0.344437   \n",
206 |        "3               0.000000                    0.000000                 0.344437   \n",
207 |        "4               0.000000                    0.000000                 0.344437   \n",
208 |        "\n",
209 |        "   LOCAL_BELONEYR_FF_MON_AVG_BAL  LOCAL_OVEONEYR_FF_SLOPE  LOCAL_SAV_SLOPE  \\\n",
210 |        "0                            0.0                 0.333284         0.333196   \n",
211 |        "1                            0.0                 0.293169         0.345716   \n",
212 |        "2                            0.0                 0.344442         0.344545   \n",
213 |        "3                            0.0                 0.344442         0.154644   \n",
214 |        "4                            0.0                 0.344442         0.344450   \n",
215 |        "\n",
216 |        "   LOCAL_SAV_CUR_ALL_BAL     ...       kid6_10_nan  kid11_15_U  kid11_15_Y  \\\n",
217 |        "0           9.277653e-04     ...               0.0         1.0         0.0   \n",
218 |        "1           7.225051e-04     ...               0.0         1.0         0.0   \n",
219 |        "2           4.032385e-07     ...               0.0         1.0         0.0   \n",
220 |        "3           7.529284e-04     ...               0.0         1.0         0.0   \n",
221 |        "4           7.852062e-07     ...               0.0         1.0         0.0   \n",
222 |        "\n",
223 |        "   kid11_15_nan  kid16_17_U  kid16_17_Y  kid16_17_nan  car_buy_New  \\\n",
224 |        "0           0.0         1.0         0.0           0.0          1.0   \n",
225 |        "1           0.0         1.0         0.0           0.0          1.0   \n",
226 |        "2           0.0         1.0         0.0           0.0          1.0   \n",
227 |        "3           0.0         1.0         0.0           0.0          0.0   \n",
228 |        "4           0.0         0.0         1.0           0.0          1.0   \n",
229 |        "\n",
230 |        "   car_buy_UNKNOWN  car_buy_nan  \n",
231 |        "0              0.0          0.0  \n",
232 |        "1              0.0          0.0  \n",
233 |        "2              0.0          0.0  \n",
234 |        "3              1.0          0.0  \n",
235 |        "4              0.0          0.0  \n",
236 |        "\n",
237 |        "[5 rows x 178 columns]"
238 |       ]
239 |      },
240 |      "execution_count": 2,
241 |      "metadata": {},
242 |      "output_type": "execute_result"
243 |     }
244 |    ],
245 |    "source": [
246 |     "model_data = pd.read_csv('./model_data.csv', encoding='utf-8')\n",
247 |     "model_data.head()"
248 |    ]
249 |   },
250 |   {
251 |    "cell_type": "code",
252 |    "execution_count": 3,
253 |    "metadata": {
254 |     "collapsed": true
255 |    },
256 |    "outputs": [],
257 |    "source": [
258 |     "all_features = list(model_data.columns)\n",
259 |     "all_features.remove('CHURN_CUST_IND')"
260 |    ]
261 |   },
262 |   {
263 |    "cell_type": "code",
264 |    "execution_count": 4,
265 |    "metadata": {
266 |     "collapsed": true
267 |    },
268 |    "outputs": [],
269 |    "source": [
270 |     "X_train,  X_test, y_train, y_test = train_test_split(model_data[all_features], model_data['CHURN_CUST_IND'], test_size=0.3, random_state=1)"
271 |    ]
272 |   },
273 |   {
274 |    "cell_type": "code",
275 |    "execution_count": 5,
276 |    "metadata": {},
277 |    "outputs": [
278 |     {
279 |      "name": "stdout",
280 |      "output_type": "stream",
281 |      "text": [
282 |       "y_train_count:  0.0    10851\n",
283 |       "1.0     1217\n",
284 |       "Name: CHURN_CUST_IND, dtype: int64\n",
285 |       "y_test_count:  0.0    4649\n",
286 |       "1.0     524\n",
287 |       "Name: CHURN_CUST_IND, dtype: int64\n"
288 |      ]
289 |     }
290 |    ],
291 |    "source": [
292 |     "print('y_train_count: ', y_train.value_counts())\n",
293 |     "print('y_test_count: ', y_test.value_counts())"
294 |    ]
295 |   },
296 |   {
297 |    "cell_type": "code",
298 |    "execution_count": 6,
299 |    "metadata": {},
300 |    "outputs": [
301 |     {
302 |      "name": "stdout",
303 |      "output_type": "stream",
304 |      "text": [
305 |       "Accuracy : 0.9115\n",
306 |       "AUC Score (Testing): 0.855843\n"
307 |      ]
308 |     }
309 |    ],
310 |    "source": [
311 |     "# 使用GBDT模块的默认参数进行训练\n",
312 |     "gbm0 = GradientBoostingClassifier(random_state=10)\n",
313 |     "gbm0.fit(X_train,y_train)\n",
314 |     "y_pred = gbm0.predict(X_test)\n",
315 |     "y_predprob = gbm0.predict_proba(X_test)[:,1]\n",
316 |     "print(\"Accuracy : %.4g\" % metrics.accuracy_score(y_test, y_pred))\n",
317 |     "print (\"AUC Score (Testing): %f\" % metrics.roc_auc_score(y_test, y_predprob))"
318 |    ]
319 |   },
320 |   {
321 |    "cell_type": "code",
322 |    "execution_count": 12,
323 |    "metadata": {},
324 |    "outputs": [
325 |     {
326 |      "data": {
327 |       "text/plain": [
328 |        "<bound method BaseEstimator.get_params of GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
329 |        "              learning_rate=0.1, loss='deviance', max_depth=3,\n",
330 |        "              max_features=None, max_leaf_nodes=None,\n",
331 |        "              min_impurity_decrease=0.0, min_impurity_split=None,\n",
332 |        "              min_samples_leaf=1, min_samples_split=2,\n",
333 |        "              min_weight_fraction_leaf=0.0, n_estimators=100,\n",
334 |        "              presort='auto', random_state=10, subsample=1.0, verbose=0,\n",
335 |        "              warm_start=False)>"
336 |       ]
337 |      },
338 |      "execution_count": 12,
339 |      "metadata": {},
340 |      "output_type": "execute_result"
341 |     }
342 |    ],
343 |    "source": [
344 |     "gbm0.get_params"
345 |    ]
346 |   },
347 |   {
348 |    "cell_type": "code",
349 |    "execution_count": 20,
350 |    "metadata": {},
351 |    "outputs": [
352 |     {
353 |      "data": {
354 |       "text/plain": [
355 |        "GridSearchCV(cv=5, error_score='raise',\n",
356 |        "       estimator=GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
357 |        "              learning_rate=0.1, loss='deviance', max_depth=5,\n",
358 |        "              max_features=1.0, max_leaf_nodes=None,\n",
359 |        "              min_impurity_decrease=0.0, min_impurity_split=None,\n",
360 |        "              min_samples_leaf=20, min_samples_split=300,\n",
361 |        "              min_weight_fraction_leaf=0.0, n_estimators=100,\n",
362 |        "              presort='auto', random_state=10, subsample=0.8, verbose=0,\n",
363 |        "              warm_start=False),\n",
364 |        "       fit_params=None, iid=False, n_jobs=1,\n",
365 |        "       param_grid={'n_estimators': range(50, 201, 10)},\n",
366 |        "       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',\n",
367 |        "       scoring='roc_auc', verbose=0)"
368 |       ]
369 |      },
370 |      "execution_count": 20,
371 |      "metadata": {},
372 |      "output_type": "execute_result"
373 |     }
374 |    ],
375 |    "source": [
376 |     "# 对estimator进行参数调优\n",
377 |     "param_test1 = {'n_estimators': range(50, 201, 10)}\n",
378 |     "est1 = GridSearchCV(estimator=GradientBoostingClassifier(learning_rate=0.1, min_samples_split=300, min_samples_leaf=20, max_depth=5, \n",
379 |     "                                                                max_features=1.0, subsample=0.8, random_state=10), \n",
380 |     "                           param_grid = param_test1, scoring='roc_auc', iid=False, cv=5)\n",
381 |     "est1.fit(X_train, y_train)"
382 |    ]
383 |   },
384 |   {
385 |    "cell_type": "code",
386 |    "execution_count": 21,
387 |    "metadata": {},
388 |    "outputs": [
389 |     {
390 |      "name": "stderr",
391 |      "output_type": "stream",
392 |      "text": [
393 |       "D:\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_search.py:761: DeprecationWarning: The grid_scores_ attribute was deprecated in version 0.18 in favor of the more elaborate cv_results_ attribute. The grid_scores_ attribute will not be available from 0.20\n",
394 |       "  DeprecationWarning)\n"
395 |      ]
396 |     },
397 |     {
398 |      "data": {
399 |       "text/plain": [
400 |        "[mean: 0.85970, std: 0.01218, params: {'n_estimators': 50},\n",
401 |        " mean: 0.85998, std: 0.01228, params: {'n_estimators': 60},\n",
402 |        " mean: 0.86043, std: 0.01162, params: {'n_estimators': 70},\n",
403 |        " mean: 0.86064, std: 0.01188, params: {'n_estimators': 80},\n",
404 |        " mean: 0.86067, std: 0.01115, params: {'n_estimators': 90},\n",
405 |        " mean: 0.86028, std: 0.01136, params: {'n_estimators': 100},\n",
406 |        " mean: 0.85983, std: 0.01133, params: {'n_estimators': 110},\n",
407 |        " mean: 0.85983, std: 0.01136, params: {'n_estimators': 120},\n",
408 |        " mean: 0.85962, std: 0.01143, params: {'n_estimators': 130},\n",
409 |        " mean: 0.85910, std: 0.01211, params: {'n_estimators': 140},\n",
410 |        " mean: 0.85833, std: 0.01207, params: {'n_estimators': 150},\n",
411 |        " mean: 0.85808, std: 0.01135, params: {'n_estimators': 160},\n",
412 |        " mean: 0.85829, std: 0.01167, params: {'n_estimators': 170},\n",
413 |        " mean: 0.85774, std: 0.01135, params: {'n_estimators': 180},\n",
414 |        " mean: 0.85699, std: 0.01170, params: {'n_estimators': 190},\n",
415 |        " mean: 0.85577, std: 0.01126, params: {'n_estimators': 200}]"
416 |       ]
417 |      },
418 |      "execution_count": 21,
419 |      "metadata": {},
420 |      "output_type": "execute_result"
421 |     }
422 |    ],
423 |    "source": [
424 |     "est1.grid_scores_"
425 |    ]
426 |   },
427 |   {
428 |    "cell_type": "code",
429 |    "execution_count": 22,
430 |    "metadata": {},
431 |    "outputs": [
432 |     {
433 |      "data": {
434 |       "text/plain": [
435 |        "{'n_estimators': 90}"
436 |       ]
437 |      },
438 |      "execution_count": 22,
439 |      "metadata": {},
440 |      "output_type": "execute_result"
441 |     }
442 |    ],
443 |    "source": [
444 |     "est1.best_params_"
445 |    ]
446 |   },
447 |   {
448 |    "cell_type": "code",
449 |    "execution_count": 23,
450 |    "metadata": {},
451 |    "outputs": [
452 |     {
453 |      "data": {
454 |       "text/plain": [
455 |        "0.8606705972896014"
456 |       ]
457 |      },
458 |      "execution_count": 23,
459 |      "metadata": {},
460 |      "output_type": "execute_result"
461 |     }
462 |    ],
463 |    "source": [
464 |     "est1.best_score_"
465 |    ]
466 |   },
467 |   {
468 |    "cell_type": "code",
469 |    "execution_count": 25,
470 |    "metadata": {},
471 |    "outputs": [
472 |     {
473 |      "name": "stdout",
474 |      "output_type": "stream",
475 |      "text": [
476 |       "Accuracy : 0.9103\n",
477 |       "AUC Score (Testing): 0.858071\n"
478 |      ]
479 |     }
480 |    ],
481 |    "source": [
482 |     "gbm1 = GradientBoostingClassifier(n_estimators=90, learning_rate=0.1, min_samples_split=300, min_samples_leaf=20, max_depth=8, \n",
483 |     "                                max_features=1.0, subsample=0.8, random_state=10)\n",
484 |     "gbm1.fit(X_train,y_train)\n",
485 |     "y_pred = gbm1.predict(X_test)\n",
486 |     "y_predprob = gbm1.predict_proba(X_test)[:,1]\n",
487 |     "print(\"Accuracy : %.4g\" % metrics.accuracy_score(y_test, y_pred))\n",
488 |     "print (\"AUC Score (Testing): %f\" % metrics.roc_auc_score(y_test, y_predprob))"
489 |    ]
490 |   },
491 |   {
492 |    "cell_type": "code",
493 |    "execution_count": 31,
494 |    "metadata": {},
495 |    "outputs": [
496 |     {
497 |      "data": {
498 |       "text/plain": [
499 |        "GridSearchCV(cv=5, error_score='raise',\n",
500 |        "       estimator=GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
501 |        "              learning_rate=0.1, loss='deviance', max_depth=3,\n",
502 |        "              max_features=1.0, max_leaf_nodes=None,\n",
503 |        "              min_impurity_decrease=0.0, min_impurity_split=None,\n",
504 |        "              min_samples_leaf=20, min_samples_split=300,\n",
505 |        "              min_weight_fraction_leaf=0.0, n_estimators=90,\n",
506 |        "              presort='auto', random_state=10, subsample=0.8, verbose=0,\n",
507 |        "              warm_start=False),\n",
508 |        "       fit_params=None, iid=False, n_jobs=1,\n",
509 |        "       param_grid={'max_depth': range(5, 16, 2)}, pre_dispatch='2*n_jobs',\n",
510 |        "       refit=True, return_train_score='warn', scoring='roc_auc', verbose=0)"
511 |       ]
512 |      },
513 |      "execution_count": 31,
514 |      "metadata": {},
515 |      "output_type": "execute_result"
516 |     }
517 |    ],
518 |    "source": [
519 |     "# 对max_depth进行调优\n",
520 |     "param_test2 = {'max_depth': range(5, 16, 2)}\n",
521 |     "est2 = GridSearchCV(estimator=GradientBoostingClassifier(n_estimators=90, learning_rate=0.1, min_samples_split=300, min_samples_leaf=20,\n",
522 |     "                                max_features=1.0, subsample=0.8, random_state=10),\n",
523 |     "                   param_grid=param_test2, scoring='roc_auc', iid=False, cv=5)\n",
524 |     "est2.fit(X_train, y_train)"
525 |    ]
526 |   },
527 |   {
528 |    "cell_type": "code",
529 |    "execution_count": 32,
530 |    "metadata": {},
531 |    "outputs": [
532 |     {
533 |      "name": "stderr",
534 |      "output_type": "stream",
535 |      "text": [
536 |       "D:\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_search.py:761: DeprecationWarning: The grid_scores_ attribute was deprecated in version 0.18 in favor of the more elaborate cv_results_ attribute. The grid_scores_ attribute will not be available from 0.20\n",
537 |       "  DeprecationWarning)\n"
538 |      ]
539 |     },
540 |     {
541 |      "data": {
542 |       "text/plain": [
543 |        "[mean: 0.86067, std: 0.01115, params: {'max_depth': 5},\n",
544 |        " mean: 0.85743, std: 0.00934, params: {'max_depth': 7},\n",
545 |        " mean: 0.85636, std: 0.01054, params: {'max_depth': 9},\n",
546 |        " mean: 0.85099, std: 0.00972, params: {'max_depth': 11},\n",
547 |        " mean: 0.85169, std: 0.01273, params: {'max_depth': 13},\n",
548 |        " mean: 0.84906, std: 0.01225, params: {'max_depth': 15}]"
549 |       ]
550 |      },
551 |      "execution_count": 32,
552 |      "metadata": {},
553 |      "output_type": "execute_result"
554 |     }
555 |    ],
556 |    "source": [
557 |     "est2.grid_scores_"
558 |    ]
559 |   },
560 |   {
561 |    "cell_type": "code",
562 |    "execution_count": 33,
563 |    "metadata": {},
564 |    "outputs": [
565 |     {
566 |      "data": {
567 |       "text/plain": [
568 |        "{'max_depth': 5}"
569 |       ]
570 |      },
571 |      "execution_count": 33,
572 |      "metadata": {},
573 |      "output_type": "execute_result"
574 |     }
575 |    ],
576 |    "source": [
577 |     "est2.best_params_"
578 |    ]
579 |   },
580 |   {
581 |    "cell_type": "code",
582 |    "execution_count": 34,
583 |    "metadata": {},
584 |    "outputs": [
585 |     {
586 |      "data": {
587 |       "text/plain": [
588 |        "0.8606705972896014"
589 |       ]
590 |      },
591 |      "execution_count": 34,
592 |      "metadata": {},
593 |      "output_type": "execute_result"
594 |     }
595 |    ],
596 |    "source": [
597 |     "est2.best_score_"
598 |    ]
599 |   },
600 |   {
601 |    "cell_type": "code",
602 |    "execution_count": 35,
603 |    "metadata": {},
604 |    "outputs": [
605 |     {
606 |      "name": "stdout",
607 |      "output_type": "stream",
608 |      "text": [
609 |       "Accuracy : 0.9126\n",
610 |       "AUC Score (Testing): 0.859257\n"
611 |      ]
612 |     }
613 |    ],
614 |    "source": [
615 |     "gbm2 = GradientBoostingClassifier(n_estimators=90, learning_rate=0.1, min_samples_split=300, min_samples_leaf=20, max_depth=5, \n",
616 |     "                                max_features=1.0, subsample=0.8, random_state=10)\n",
617 |     "gbm2.fit(X_train,y_train)\n",
618 |     "y_pred = gbm2.predict(X_test)\n",
619 |     "y_predprob = gbm2.predict_proba(X_test)[:,1]\n",
620 |     "print(\"Accuracy : %.4g\" % metrics.accuracy_score(y_test, y_pred))\n",
621 |     "print (\"AUC Score (Testing): %f\" % metrics.roc_auc_score(y_test, y_predprob))"
622 |    ]
623 |   },
624 |   {
625 |    "cell_type": "code",
626 |    "execution_count": 37,
627 |    "metadata": {},
628 |    "outputs": [
629 |     {
630 |      "data": {
631 |       "text/plain": [
632 |        "GridSearchCV(cv=5, error_score='raise',\n",
633 |        "       estimator=GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
634 |        "              learning_rate=0.1, loss='deviance', max_depth=5,\n",
635 |        "              max_features=1.0, max_leaf_nodes=None,\n",
636 |        "              min_impurity_decrease=0.0, min_impurity_split=None,\n",
637 |        "              min_samples_leaf=20, min_samples_split=300,\n",
638 |        "              min_weight_fraction_leaf=0.0, n_estimators=90,\n",
639 |        "              presort='auto', random_state=10, subsample=0.8, verbose=0,\n",
640 |        "              warm_start=False),\n",
641 |        "       fit_params=None, iid=False, n_jobs=1,\n",
642 |        "       param_grid={'learning_rate': [0.01, 0.05, 0.1, 0.15]},\n",
643 |        "       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',\n",
644 |        "       scoring='roc_auc', verbose=0)"
645 |       ]
646 |      },
647 |      "execution_count": 37,
648 |      "metadata": {},
649 |      "output_type": "execute_result"
650 |     }
651 |    ],
652 |    "source": [
653 |     "# 对learning_rate进行调优\n",
654 |     "param_test3 = {'learning_rate': [0.01, 0.05, 0.1, 0.15]}\n",
655 |     "est3 = GridSearchCV(estimator=GradientBoostingClassifier(n_estimators=90, max_depth=5, min_samples_split=300, min_samples_leaf=20,\n",
656 |     "                                max_features=1.0, subsample=0.8, random_state=10),\n",
657 |     "                   param_grid=param_test3, scoring='roc_auc', iid=False, cv=5)\n",
658 |     "est3.fit(X_train, y_train)"
659 |    ]
660 |   },
661 |   {
662 |    "cell_type": "code",
663 |    "execution_count": 38,
664 |    "metadata": {},
665 |    "outputs": [
666 |     {
667 |      "name": "stderr",
668 |      "output_type": "stream",
669 |      "text": [
670 |       "D:\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_search.py:761: DeprecationWarning: The grid_scores_ attribute was deprecated in version 0.18 in favor of the more elaborate cv_results_ attribute. The grid_scores_ attribute will not be available from 0.20\n",
671 |       "  DeprecationWarning)\n"
672 |      ]
673 |     },
674 |     {
675 |      "data": {
676 |       "text/plain": [
677 |        "[mean: 0.85457, std: 0.01371, params: {'learning_rate': 0.01},\n",
678 |        " mean: 0.86058, std: 0.01204, params: {'learning_rate': 0.05},\n",
679 |        " mean: 0.86067, std: 0.01115, params: {'learning_rate': 0.1},\n",
680 |        " mean: 0.85709, std: 0.01315, params: {'learning_rate': 0.15}]"
681 |       ]
682 |      },
683 |      "execution_count": 38,
684 |      "metadata": {},
685 |      "output_type": "execute_result"
686 |     }
687 |    ],
688 |    "source": [
689 |     "est3.grid_scores_"
690 |    ]
691 |   },
692 |   {
693 |    "cell_type": "code",
694 |    "execution_count": 39,
695 |    "metadata": {},
696 |    "outputs": [
697 |     {
698 |      "data": {
699 |       "text/plain": [
700 |        "{'learning_rate': 0.1}"
701 |       ]
702 |      },
703 |      "execution_count": 39,
704 |      "metadata": {},
705 |      "output_type": "execute_result"
706 |     }
707 |    ],
708 |    "source": [
709 |     "est3.best_params_"
710 |    ]
711 |   },
712 |   {
713 |    "cell_type": "code",
714 |    "execution_count": 40,
715 |    "metadata": {},
716 |    "outputs": [
717 |     {
718 |      "data": {
719 |       "text/plain": [
720 |        "0.8606705972896014"
721 |       ]
722 |      },
723 |      "execution_count": 40,
724 |      "metadata": {},
725 |      "output_type": "execute_result"
726 |     }
727 |    ],
728 |    "source": [
729 |     "est3.best_score_"
730 |    ]
731 |   },
732 |   {
733 |    "cell_type": "code",
734 |    "execution_count": 41,
735 |    "metadata": {},
736 |    "outputs": [
737 |     {
738 |      "data": {
739 |       "text/plain": [
740 |        "GridSearchCV(cv=5, error_score='raise',\n",
741 |        "       estimator=GradientBoostingClassifier(criterion='friedman_mse', init=None,\n",
742 |        "              learning_rate=0.1, loss='deviance', max_depth=5,\n",
743 |        "              max_features=None, max_leaf_nodes=None,\n",
744 |        "              min_impurity_decrease=0.0, min_impurity_split=None,\n",
745 |        "              min_samples_leaf=20, min_samples_split=300,\n",
746 |        "              min_weight_fraction_leaf=0.0, n_estimators=90,\n",
747 |        "              presort='auto', random_state=10, subsample=0.8, verbose=0,\n",
748 |        "              warm_start=False),\n",
749 |        "       fit_params=None, iid=False, n_jobs=1,\n",
750 |        "       param_grid={'max_features': (0.3, 0.5, 0.8, 1.0)},\n",
751 |        "       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',\n",
752 |        "       scoring='roc_auc', verbose=0)"
753 |       ]
754 |      },
755 |      "execution_count": 41,
756 |      "metadata": {},
757 |      "output_type": "execute_result"
758 |     }
759 |    ],
760 |    "source": [
761 |     "# 对max_features进行调优\n",
762 |     "param_test4 = {'max_features': (0.3, 0.5, 0.8, 1.0)}\n",
763 |     "est4 = GridSearchCV(estimator=GradientBoostingClassifier(n_estimators=90, learning_rate=0.1, min_samples_split=300, min_samples_leaf=20,\n",
764 |     "                                max_depth=5, subsample=0.8, random_state=10),\n",
765 |     "                   param_grid=param_test4, scoring='roc_auc', iid=False, cv=5)\n",
766 |     "est4.fit(X_train, y_train)"
767 |    ]
768 |   },
769 |   {
770 |    "cell_type": "code",
771 |    "execution_count": 42,
772 |    "metadata": {},
773 |    "outputs": [
774 |     {
775 |      "name": "stderr",
776 |      "output_type": "stream",
777 |      "text": [
778 |       "D:\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_search.py:761: DeprecationWarning: The grid_scores_ attribute was deprecated in version 0.18 in favor of the more elaborate cv_results_ attribute. The grid_scores_ attribute will not be available from 0.20\n",
779 |       "  DeprecationWarning)\n"
780 |      ]
781 |     },
782 |     {
783 |      "data": {
784 |       "text/plain": [
785 |        "[mean: 0.85937, std: 0.01201, params: {'max_features': 0.3},\n",
786 |        " mean: 0.85815, std: 0.01321, params: {'max_features': 0.5},\n",
787 |        " mean: 0.85989, std: 0.01139, params: {'max_features': 0.8},\n",
788 |        " mean: 0.86067, std: 0.01115, params: {'max_features': 1.0}]"
789 |       ]
790 |      },
791 |      "execution_count": 42,
792 |      "metadata": {},
793 |      "output_type": "execute_result"
794 |     }
795 |    ],
796 |    "source": [
797 |     "est4.grid_scores_"
798 |    ]
799 |   },
800 |   {
801 |    "cell_type": "code",
802 |    "execution_count": 43,
803 |    "metadata": {},
804 |    "outputs": [
805 |     {
806 |      "data": {
807 |       "text/plain": [
808 |        "{'max_features': 1.0}"
809 |       ]
810 |      },
811 |      "execution_count": 43,
812 |      "metadata": {},
813 |      "output_type": "execute_result"
814 |     }
815 |    ],
816 |    "source": [
817 |     "est4.best_params_"
818 |    ]
819 |   },
820 |   {
821 |    "cell_type": "code",
822 |    "execution_count": 44,
823 |    "metadata": {},
824 |    "outputs": [
825 |     {
826 |      "data": {
827 |       "text/plain": [
828 |        "0.8606705972896014"
829 |       ]
830 |      },
831 |      "execution_count": 44,
832 |      "metadata": {},
833 |      "output_type": "execute_result"
834 |     }
835 |    ],
836 |    "source": [
837 |     "est4.best_score_"
838 |    ]
839 |   },
840 |   {
841 |    "cell_type": "code",
842 |    "execution_count": 51,
843 |    "metadata": {},
844 |    "outputs": [
845 |     {
846 |      "data": {
847 |       "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAGDCAYAAADEegxVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
848 |       "text/plain": [
849 |        "<matplotlib.figure.Figure at 0x1b6cdf2f2b0>"
850 |       ]
851 |      },
852 |      "metadata": {},
853 |      "output_type": "display_data"
854 |     }
855 |    ],
856 |    "source": [
857 |     "# 综合以上的参数调优，重新构建一个新的模型\n",
858 |     "clf = GradientBoostingClassifier(n_estimators=90, learning_rate=0.1, min_samples_split=300, min_samples_leaf=20,\n",
859 |     "                                max_depth=5, max_features=1.0, subsample=0.8, random_state=10)\n",
860 |     "clf.fit(X_train, y_train)\n",
861 |     "y_predict_proba = clf.predict_proba(X_test)[:,1]\n",
862 |     "fpr, tpr, thresholds = roc_curve(y_test, y_predict_proba)\n",
863 |     "roc_auc = auc(fpr, tpr)\n",
864 |     "fig = plt.figure(figsize=(6, 6))\n",
865 |     "ax = fig.add_subplot(111)\n",
866 |     "ax.plot(fpr, tpr, lw=1, label='ROC (area = %.2f)' % (roc_auc))\n",
867 |     "ax.plot([0, 1], [0, 1], '--', color='red', label='luck')\n",
868 |     "ax.set_xlabel('False Positive Rate')\n",
869 |     "ax.set_ylabel('True Positive Rate')\n",
870 |     "ax.set_title('Receive Operating Characteristic')\n",
871 |     "ax.legend(loc='lower right')\n",
872 |     "plt.show()"
873 |    ]
874 |   },
875 |   {
876 |    "cell_type": "code",
877 |    "execution_count": null,
878 |    "metadata": {
879 |     "collapsed": true
880 |    },
881 |    "outputs": [],
882 |    "source": []
883 |   }
884 |  ],
885 |  "metadata": {
886 |   "kernelspec": {
887 |    "display_name": "Python 3",
888 |    "language": "python",
889 |    "name": "python3"
890 |   },
891 |   "language_info": {
892 |    "codemirror_mode": {
893 |     "name": "ipython",
894 |     "version": 3
895 |    },
896 |    "file_extension": ".py",
897 |    "mimetype": "text/x-python",
898 |    "name": "python",
899 |    "nbconvert_exporter": "python",
900 |    "pygments_lexer": "ipython3",
901 |    "version": "3.6.3"
902 |   }
903 |  },
904 |  "nbformat": 4,
905 |  "nbformat_minor": 2
906 | }
907 | 


--------------------------------------------------------------------------------
/bank churn model - preprocessing.ipynb:
--------------------------------------------------------------------------------
   1 | {
   2 |  "cells": [
   3 |   {
   4 |    "cell_type": "code",
   5 |    "execution_count": 1,
   6 |    "metadata": {
   7 |     "collapsed": true
   8 |    },
   9 |    "outputs": [],
  10 |    "source": [
  11 |     "import numpy as np\n",
  12 |     "import pandas as pd\n",
  13 |     "import random\n",
  14 |     "import operator\n",
  15 |     "import numbers"
  16 |    ]
  17 |   },
  18 |   {
  19 |    "cell_type": "code",
  20 |    "execution_count": 2,
  21 |    "metadata": {},
  22 |    "outputs": [
  23 |     {
  24 |      "data": {
  25 |       "text/html": [
  26 |        "<div>\n",
  27 |        "<style>\n",
  28 |        "    .dataframe thead tr:only-child th {\n",
  29 |        "        text-align: right;\n",
  30 |        "    }\n",
  31 |        "\n",
  32 |        "    .dataframe thead th {\n",
  33 |        "        text-align: left;\n",
  34 |        "    }\n",
  35 |        "\n",
  36 |        "    .dataframe tbody tr th {\n",
  37 |        "        vertical-align: top;\n",
  38 |        "    }\n",
  39 |        "</style>\n",
  40 |        "<table border=\"1\" class=\"dataframe\">\n",
  41 |        "  <thead>\n",
  42 |        "    <tr style=\"text-align: right;\">\n",
  43 |        "      <th></th>\n",
  44 |        "      <th>CUST_ID</th>\n",
  45 |        "      <th>OPEN_ACC_DUR</th>\n",
  46 |        "      <th>AGE</th>\n",
  47 |        "      <th>GENDER_CD</th>\n",
  48 |        "      <th>HASNT_HOME_ADDRESS_INF</th>\n",
  49 |        "      <th>HASNT_MOBILE_TEL_NUM_INF</th>\n",
  50 |        "      <th>LOCAL_CUR_SAV_SLOPE</th>\n",
  51 |        "      <th>LOCAL_CUR_MON_AVG_BAL</th>\n",
  52 |        "      <th>LOCAL_CUR_MON_AVG_BAL_PROP</th>\n",
  53 |        "      <th>LOCAL_CUR_ACCT_NUM</th>\n",
  54 |        "      <th>...</th>\n",
  55 |        "      <th>educ1</th>\n",
  56 |        "      <th>proptype</th>\n",
  57 |        "      <th>pcowner</th>\n",
  58 |        "      <th>ethnic</th>\n",
  59 |        "      <th>kid0_2</th>\n",
  60 |        "      <th>kid3_5</th>\n",
  61 |        "      <th>kid6_10</th>\n",
  62 |        "      <th>kid11_15</th>\n",
  63 |        "      <th>kid16_17</th>\n",
  64 |        "      <th>car_buy</th>\n",
  65 |        "    </tr>\n",
  66 |        "  </thead>\n",
  67 |        "  <tbody>\n",
  68 |        "    <tr>\n",
  69 |        "      <th>0</th>\n",
  70 |        "      <td>1</td>\n",
  71 |        "      <td>231</td>\n",
  72 |        "      <td>82</td>\n",
  73 |        "      <td>1</td>\n",
  74 |        "      <td>Y</td>\n",
  75 |        "      <td>N</td>\n",
  76 |        "      <td>0.000000</td>\n",
  77 |        "      <td>0.00</td>\n",
  78 |        "      <td>0.000000</td>\n",
  79 |        "      <td>0</td>\n",
  80 |        "      <td>...</td>\n",
  81 |        "      <td>2.0</td>\n",
  82 |        "      <td>NaN</td>\n",
  83 |        "      <td>Y</td>\n",
  84 |        "      <td>S</td>\n",
  85 |        "      <td>U</td>\n",
  86 |        "      <td>U</td>\n",
  87 |        "      <td>U</td>\n",
  88 |        "      <td>U</td>\n",
  89 |        "      <td>U</td>\n",
  90 |        "      <td>New</td>\n",
  91 |        "    </tr>\n",
  92 |        "    <tr>\n",
  93 |        "      <th>1</th>\n",
  94 |        "      <td>2</td>\n",
  95 |        "      <td>48</td>\n",
  96 |        "      <td>42</td>\n",
  97 |        "      <td>1</td>\n",
  98 |        "      <td>Y</td>\n",
  99 |        "      <td>N</td>\n",
 100 |        "      <td>1.005692</td>\n",
 101 |        "      <td>20149.04</td>\n",
 102 |        "      <td>0.264435</td>\n",
 103 |        "      <td>3</td>\n",
 104 |        "      <td>...</td>\n",
 105 |        "      <td>NaN</td>\n",
 106 |        "      <td>NaN</td>\n",
 107 |        "      <td>NaN</td>\n",
 108 |        "      <td>N</td>\n",
 109 |        "      <td>U</td>\n",
 110 |        "      <td>U</td>\n",
 111 |        "      <td>U</td>\n",
 112 |        "      <td>U</td>\n",
 113 |        "      <td>U</td>\n",
 114 |        "      <td>New</td>\n",
 115 |        "    </tr>\n",
 116 |        "    <tr>\n",
 117 |        "      <th>2</th>\n",
 118 |        "      <td>3</td>\n",
 119 |        "      <td>102</td>\n",
 120 |        "      <td>31</td>\n",
 121 |        "      <td>2</td>\n",
 122 |        "      <td>Y</td>\n",
 123 |        "      <td>N</td>\n",
 124 |        "      <td>0.000562</td>\n",
 125 |        "      <td>17.81</td>\n",
 126 |        "      <td>1.000000</td>\n",
 127 |        "      <td>1</td>\n",
 128 |        "      <td>...</td>\n",
 129 |        "      <td>NaN</td>\n",
 130 |        "      <td>A</td>\n",
 131 |        "      <td>Y</td>\n",
 132 |        "      <td>F</td>\n",
 133 |        "      <td>U</td>\n",
 134 |        "      <td>U</td>\n",
 135 |        "      <td>Y</td>\n",
 136 |        "      <td>U</td>\n",
 137 |        "      <td>U</td>\n",
 138 |        "      <td>New</td>\n",
 139 |        "    </tr>\n",
 140 |        "    <tr>\n",
 141 |        "      <th>3</th>\n",
 142 |        "      <td>4</td>\n",
 143 |        "      <td>62</td>\n",
 144 |        "      <td>78</td>\n",
 145 |        "      <td>2</td>\n",
 146 |        "      <td>Y</td>\n",
 147 |        "      <td>N</td>\n",
 148 |        "      <td>-0.550912</td>\n",
 149 |        "      <td>29359.21</td>\n",
 150 |        "      <td>1.000000</td>\n",
 151 |        "      <td>1</td>\n",
 152 |        "      <td>...</td>\n",
 153 |        "      <td>NaN</td>\n",
 154 |        "      <td>NaN</td>\n",
 155 |        "      <td>NaN</td>\n",
 156 |        "      <td>N</td>\n",
 157 |        "      <td>U</td>\n",
 158 |        "      <td>U</td>\n",
 159 |        "      <td>U</td>\n",
 160 |        "      <td>U</td>\n",
 161 |        "      <td>U</td>\n",
 162 |        "      <td>UNKNOWN</td>\n",
 163 |        "    </tr>\n",
 164 |        "    <tr>\n",
 165 |        "      <th>4</th>\n",
 166 |        "      <td>5</td>\n",
 167 |        "      <td>79</td>\n",
 168 |        "      <td>46</td>\n",
 169 |        "      <td>1</td>\n",
 170 |        "      <td>Y</td>\n",
 171 |        "      <td>N</td>\n",
 172 |        "      <td>0.000288</td>\n",
 173 |        "      <td>34.68</td>\n",
 174 |        "      <td>1.000000</td>\n",
 175 |        "      <td>1</td>\n",
 176 |        "      <td>...</td>\n",
 177 |        "      <td>NaN</td>\n",
 178 |        "      <td>A</td>\n",
 179 |        "      <td>Y</td>\n",
 180 |        "      <td>U</td>\n",
 181 |        "      <td>U</td>\n",
 182 |        "      <td>U</td>\n",
 183 |        "      <td>U</td>\n",
 184 |        "      <td>U</td>\n",
 185 |        "      <td>Y</td>\n",
 186 |        "      <td>New</td>\n",
 187 |        "    </tr>\n",
 188 |        "  </tbody>\n",
 189 |        "</table>\n",
 190 |        "<p>5 rows × 127 columns</p>\n",
 191 |        "</div>"
 192 |       ],
 193 |       "text/plain": [
 194 |        "   CUST_ID  OPEN_ACC_DUR  AGE GENDER_CD HASNT_HOME_ADDRESS_INF  \\\n",
 195 |        "0        1           231   82         1                      Y   \n",
 196 |        "1        2            48   42         1                      Y   \n",
 197 |        "2        3           102   31         2                      Y   \n",
 198 |        "3        4            62   78         2                      Y   \n",
 199 |        "4        5            79   46         1                      Y   \n",
 200 |        "\n",
 201 |        "  HASNT_MOBILE_TEL_NUM_INF  LOCAL_CUR_SAV_SLOPE  LOCAL_CUR_MON_AVG_BAL  \\\n",
 202 |        "0                        N             0.000000                   0.00   \n",
 203 |        "1                        N             1.005692               20149.04   \n",
 204 |        "2                        N             0.000562                  17.81   \n",
 205 |        "3                        N            -0.550912               29359.21   \n",
 206 |        "4                        N             0.000288                  34.68   \n",
 207 |        "\n",
 208 |        "   LOCAL_CUR_MON_AVG_BAL_PROP  LOCAL_CUR_ACCT_NUM   ...     educ1  proptype  \\\n",
 209 |        "0                    0.000000                   0   ...       2.0       NaN   \n",
 210 |        "1                    0.264435                   3   ...       NaN       NaN   \n",
 211 |        "2                    1.000000                   1   ...       NaN         A   \n",
 212 |        "3                    1.000000                   1   ...       NaN       NaN   \n",
 213 |        "4                    1.000000                   1   ...       NaN         A   \n",
 214 |        "\n",
 215 |        "   pcowner  ethnic  kid0_2  kid3_5  kid6_10  kid11_15  kid16_17  car_buy  \n",
 216 |        "0        Y       S       U       U        U         U         U      New  \n",
 217 |        "1      NaN       N       U       U        U         U         U      New  \n",
 218 |        "2        Y       F       U       U        Y         U         U      New  \n",
 219 |        "3      NaN       N       U       U        U         U         U  UNKNOWN  \n",
 220 |        "4        Y       U       U       U        U         U         Y      New  \n",
 221 |        "\n",
 222 |        "[5 rows x 127 columns]"
 223 |       ]
 224 |      },
 225 |      "execution_count": 2,
 226 |      "metadata": {},
 227 |      "output_type": "execute_result"
 228 |     }
 229 |    ],
 230 |    "source": [
 231 |     "bank_churn = pd.read_csv('./bankChurn.csv', encoding='utf-8')\n",
 232 |     "external_data = pd.read_csv('./ExternalData.csv', encoding='utf-8')\n",
 233 |     "df = pd.merge(bank_churn, external_data, on='CUST_ID')\n",
 234 |     "df.head()"
 235 |    ]
 236 |   },
 237 |   {
 238 |    "cell_type": "code",
 239 |    "execution_count": 3,
 240 |    "metadata": {
 241 |     "collapsed": true
 242 |    },
 243 |    "outputs": [],
 244 |    "source": [
 245 |     "model_data = df.copy()\n",
 246 |     "indep_cols = list(model_data.columns)\n",
 247 |     "indep_cols.remove('CHURN_CUST_IND')\n",
 248 |     "indep_cols.remove('CUST_ID')"
 249 |    ]
 250 |   },
 251 |   {
 252 |    "cell_type": "code",
 253 |    "execution_count": 4,
 254 |    "metadata": {},
 255 |    "outputs": [
 256 |     {
 257 |      "data": {
 258 |       "text/html": [
 259 |        "<div>\n",
 260 |        "<style>\n",
 261 |        "    .dataframe thead tr:only-child th {\n",
 262 |        "        text-align: right;\n",
 263 |        "    }\n",
 264 |        "\n",
 265 |        "    .dataframe thead th {\n",
 266 |        "        text-align: left;\n",
 267 |        "    }\n",
 268 |        "\n",
 269 |        "    .dataframe tbody tr th {\n",
 270 |        "        vertical-align: top;\n",
 271 |        "    }\n",
 272 |        "</style>\n",
 273 |        "<table border=\"1\" class=\"dataframe\">\n",
 274 |        "  <thead>\n",
 275 |        "    <tr style=\"text-align: right;\">\n",
 276 |        "      <th></th>\n",
 277 |        "      <th>count</th>\n",
 278 |        "      <th>unique</th>\n",
 279 |        "      <th>top</th>\n",
 280 |        "      <th>freq</th>\n",
 281 |        "      <th>missing_pct</th>\n",
 282 |        "    </tr>\n",
 283 |        "  </thead>\n",
 284 |        "  <tbody>\n",
 285 |        "    <tr>\n",
 286 |        "      <th>GENDER_CD</th>\n",
 287 |        "      <td>17241</td>\n",
 288 |        "      <td>3</td>\n",
 289 |        "      <td>2</td>\n",
 290 |        "      <td>8875</td>\n",
 291 |        "      <td>0.000000</td>\n",
 292 |        "    </tr>\n",
 293 |        "    <tr>\n",
 294 |        "      <th>HASNT_HOME_ADDRESS_INF</th>\n",
 295 |        "      <td>17241</td>\n",
 296 |        "      <td>2</td>\n",
 297 |        "      <td>Y</td>\n",
 298 |        "      <td>14765</td>\n",
 299 |        "      <td>0.000000</td>\n",
 300 |        "    </tr>\n",
 301 |        "    <tr>\n",
 302 |        "      <th>HASNT_MOBILE_TEL_NUM_INF</th>\n",
 303 |        "      <td>17241</td>\n",
 304 |        "      <td>2</td>\n",
 305 |        "      <td>N</td>\n",
 306 |        "      <td>11753</td>\n",
 307 |        "      <td>0.000000</td>\n",
 308 |        "    </tr>\n",
 309 |        "    <tr>\n",
 310 |        "      <th>crclscod</th>\n",
 311 |        "      <td>17241</td>\n",
 312 |        "      <td>47</td>\n",
 313 |        "      <td>AA</td>\n",
 314 |        "      <td>6287</td>\n",
 315 |        "      <td>0.000000</td>\n",
 316 |        "    </tr>\n",
 317 |        "    <tr>\n",
 318 |        "      <th>asl_flag</th>\n",
 319 |        "      <td>17241</td>\n",
 320 |        "      <td>2</td>\n",
 321 |        "      <td>N</td>\n",
 322 |        "      <td>14576</td>\n",
 323 |        "      <td>0.000000</td>\n",
 324 |        "    </tr>\n",
 325 |        "    <tr>\n",
 326 |        "      <th>last_swap</th>\n",
 327 |        "      <td>7690</td>\n",
 328 |        "      <td>980</td>\n",
 329 |        "      <td>7/5/2001</td>\n",
 330 |        "      <td>35</td>\n",
 331 |        "      <td>0.553970</td>\n",
 332 |        "    </tr>\n",
 333 |        "    <tr>\n",
 334 |        "      <th>dwlltype</th>\n",
 335 |        "      <td>11952</td>\n",
 336 |        "      <td>2</td>\n",
 337 |        "      <td>S</td>\n",
 338 |        "      <td>8595</td>\n",
 339 |        "      <td>0.306769</td>\n",
 340 |        "    </tr>\n",
 341 |        "    <tr>\n",
 342 |        "      <th>marital</th>\n",
 343 |        "      <td>16935</td>\n",
 344 |        "      <td>5</td>\n",
 345 |        "      <td>U</td>\n",
 346 |        "      <td>6237</td>\n",
 347 |        "      <td>0.017748</td>\n",
 348 |        "    </tr>\n",
 349 |        "    <tr>\n",
 350 |        "      <th>wrkwoman</th>\n",
 351 |        "      <td>2230</td>\n",
 352 |        "      <td>1</td>\n",
 353 |        "      <td>Y</td>\n",
 354 |        "      <td>2230</td>\n",
 355 |        "      <td>0.870657</td>\n",
 356 |        "    </tr>\n",
 357 |        "    <tr>\n",
 358 |        "      <th>proptype</th>\n",
 359 |        "      <td>4965</td>\n",
 360 |        "      <td>6</td>\n",
 361 |        "      <td>A</td>\n",
 362 |        "      <td>4467</td>\n",
 363 |        "      <td>0.712024</td>\n",
 364 |        "    </tr>\n",
 365 |        "    <tr>\n",
 366 |        "      <th>pcowner</th>\n",
 367 |        "      <td>3232</td>\n",
 368 |        "      <td>1</td>\n",
 369 |        "      <td>Y</td>\n",
 370 |        "      <td>3232</td>\n",
 371 |        "      <td>0.812540</td>\n",
 372 |        "    </tr>\n",
 373 |        "    <tr>\n",
 374 |        "      <th>ethnic</th>\n",
 375 |        "      <td>16935</td>\n",
 376 |        "      <td>17</td>\n",
 377 |        "      <td>N</td>\n",
 378 |        "      <td>5866</td>\n",
 379 |        "      <td>0.017748</td>\n",
 380 |        "    </tr>\n",
 381 |        "    <tr>\n",
 382 |        "      <th>kid0_2</th>\n",
 383 |        "      <td>16935</td>\n",
 384 |        "      <td>2</td>\n",
 385 |        "      <td>U</td>\n",
 386 |        "      <td>16269</td>\n",
 387 |        "      <td>0.017748</td>\n",
 388 |        "    </tr>\n",
 389 |        "    <tr>\n",
 390 |        "      <th>kid3_5</th>\n",
 391 |        "      <td>16935</td>\n",
 392 |        "      <td>2</td>\n",
 393 |        "      <td>U</td>\n",
 394 |        "      <td>16165</td>\n",
 395 |        "      <td>0.017748</td>\n",
 396 |        "    </tr>\n",
 397 |        "    <tr>\n",
 398 |        "      <th>kid6_10</th>\n",
 399 |        "      <td>16935</td>\n",
 400 |        "      <td>2</td>\n",
 401 |        "      <td>U</td>\n",
 402 |        "      <td>15500</td>\n",
 403 |        "      <td>0.017748</td>\n",
 404 |        "    </tr>\n",
 405 |        "    <tr>\n",
 406 |        "      <th>kid11_15</th>\n",
 407 |        "      <td>16935</td>\n",
 408 |        "      <td>2</td>\n",
 409 |        "      <td>U</td>\n",
 410 |        "      <td>15378</td>\n",
 411 |        "      <td>0.017748</td>\n",
 412 |        "    </tr>\n",
 413 |        "    <tr>\n",
 414 |        "      <th>kid16_17</th>\n",
 415 |        "      <td>16935</td>\n",
 416 |        "      <td>2</td>\n",
 417 |        "      <td>U</td>\n",
 418 |        "      <td>15277</td>\n",
 419 |        "      <td>0.017748</td>\n",
 420 |        "    </tr>\n",
 421 |        "    <tr>\n",
 422 |        "      <th>car_buy</th>\n",
 423 |        "      <td>16935</td>\n",
 424 |        "      <td>2</td>\n",
 425 |        "      <td>UNKNOWN</td>\n",
 426 |        "      <td>9659</td>\n",
 427 |        "      <td>0.017748</td>\n",
 428 |        "    </tr>\n",
 429 |        "  </tbody>\n",
 430 |        "</table>\n",
 431 |        "</div>"
 432 |       ],
 433 |       "text/plain": [
 434 |        "                          count unique       top   freq  missing_pct\n",
 435 |        "GENDER_CD                 17241      3         2   8875     0.000000\n",
 436 |        "HASNT_HOME_ADDRESS_INF    17241      2         Y  14765     0.000000\n",
 437 |        "HASNT_MOBILE_TEL_NUM_INF  17241      2         N  11753     0.000000\n",
 438 |        "crclscod                  17241     47        AA   6287     0.000000\n",
 439 |        "asl_flag                  17241      2         N  14576     0.000000\n",
 440 |        "last_swap                  7690    980  7/5/2001     35     0.553970\n",
 441 |        "dwlltype                  11952      2         S   8595     0.306769\n",
 442 |        "marital                   16935      5         U   6237     0.017748\n",
 443 |        "wrkwoman                   2230      1         Y   2230     0.870657\n",
 444 |        "proptype                   4965      6         A   4467     0.712024\n",
 445 |        "pcowner                    3232      1         Y   3232     0.812540\n",
 446 |        "ethnic                    16935     17         N   5866     0.017748\n",
 447 |        "kid0_2                    16935      2         U  16269     0.017748\n",
 448 |        "kid3_5                    16935      2         U  16165     0.017748\n",
 449 |        "kid6_10                   16935      2         U  15500     0.017748\n",
 450 |        "kid11_15                  16935      2         U  15378     0.017748\n",
 451 |        "kid16_17                  16935      2         U  15277     0.017748\n",
 452 |        "car_buy                   16935      2   UNKNOWN   9659     0.017748"
 453 |       ]
 454 |      },
 455 |      "execution_count": 4,
 456 |      "metadata": {},
 457 |      "output_type": "execute_result"
 458 |     }
 459 |    ],
 460 |    "source": [
 461 |     "object_df = model_data.select_dtypes(include=['object']).describe().T.assign(missing_pct = model_data.apply(lambda x: (len(x) - x.count()) / float(len(x))))\n",
 462 |     "object_df"
 463 |    ]
 464 |   },
 465 |   {
 466 |    "cell_type": "code",
 467 |    "execution_count": 5,
 468 |    "metadata": {},
 469 |    "outputs": [],
 470 |    "source": [
 471 |     "# 去除属性值过多的属性\n",
 472 |     "model_data.drop(['crclscod', 'ethnic'], axis=1, inplace=True)\n",
 473 |     "# 去除缺失值过多或无意义的属性\n",
 474 |     "model_data.drop(['last_swap'], axis=1, inplace=True)\n",
 475 |     "# wrkwoman表示是否有职场女性、proptype表示房屋类型、pcowner表示家里是否有电脑 这些属性在实际中均有很强的决策意义，以此不能删除\n",
 476 |     "# 针对这些变量可以采用对缺失值用一个哑变量表示"
 477 |    ]
 478 |   },
 479 |   {
 480 |    "cell_type": "code",
 481 |    "execution_count": 6,
 482 |    "metadata": {},
 483 |    "outputs": [
 484 |     {
 485 |      "data": {
 486 |       "text/html": [
 487 |        "<div>\n",
 488 |        "<style>\n",
 489 |        "    .dataframe thead tr:only-child th {\n",
 490 |        "        text-align: right;\n",
 491 |        "    }\n",
 492 |        "\n",
 493 |        "    .dataframe thead th {\n",
 494 |        "        text-align: left;\n",
 495 |        "    }\n",
 496 |        "\n",
 497 |        "    .dataframe tbody tr th {\n",
 498 |        "        vertical-align: top;\n",
 499 |        "    }\n",
 500 |        "</style>\n",
 501 |        "<table border=\"1\" class=\"dataframe\">\n",
 502 |        "  <thead>\n",
 503 |        "    <tr style=\"text-align: right;\">\n",
 504 |        "      <th></th>\n",
 505 |        "      <th>count</th>\n",
 506 |        "      <th>mean</th>\n",
 507 |        "      <th>std</th>\n",
 508 |        "      <th>min</th>\n",
 509 |        "      <th>25%</th>\n",
 510 |        "      <th>50%</th>\n",
 511 |        "      <th>75%</th>\n",
 512 |        "      <th>max</th>\n",
 513 |        "      <th>missing_pct</th>\n",
 514 |        "    </tr>\n",
 515 |        "  </thead>\n",
 516 |        "  <tbody>\n",
 517 |        "    <tr>\n",
 518 |        "      <th>CUST_ID</th>\n",
 519 |        "      <td>17241.0</td>\n",
 520 |        "      <td>8621.000000</td>\n",
 521 |        "      <td>4977.192331</td>\n",
 522 |        "      <td>1.0</td>\n",
 523 |        "      <td>4311.0</td>\n",
 524 |        "      <td>8621.0</td>\n",
 525 |        "      <td>12931.0</td>\n",
 526 |        "      <td>17241.0</td>\n",
 527 |        "      <td>0.0</td>\n",
 528 |        "    </tr>\n",
 529 |        "    <tr>\n",
 530 |        "      <th>OPEN_ACC_DUR</th>\n",
 531 |        "      <td>17241.0</td>\n",
 532 |        "      <td>80.104750</td>\n",
 533 |        "      <td>34.944320</td>\n",
 534 |        "      <td>0.0</td>\n",
 535 |        "      <td>56.0</td>\n",
 536 |        "      <td>75.0</td>\n",
 537 |        "      <td>104.0</td>\n",
 538 |        "      <td>278.0</td>\n",
 539 |        "      <td>0.0</td>\n",
 540 |        "    </tr>\n",
 541 |        "    <tr>\n",
 542 |        "      <th>AGE</th>\n",
 543 |        "      <td>17241.0</td>\n",
 544 |        "      <td>49.339481</td>\n",
 545 |        "      <td>15.431282</td>\n",
 546 |        "      <td>4.0</td>\n",
 547 |        "      <td>38.0</td>\n",
 548 |        "      <td>47.0</td>\n",
 549 |        "      <td>59.0</td>\n",
 550 |        "      <td>110.0</td>\n",
 551 |        "      <td>0.0</td>\n",
 552 |        "    </tr>\n",
 553 |        "    <tr>\n",
 554 |        "      <th>LOCAL_CUR_ACCT_NUM</th>\n",
 555 |        "      <td>17241.0</td>\n",
 556 |        "      <td>1.796880</td>\n",
 557 |        "      <td>3.503054</td>\n",
 558 |        "      <td>0.0</td>\n",
 559 |        "      <td>1.0</td>\n",
 560 |        "      <td>1.0</td>\n",
 561 |        "      <td>2.0</td>\n",
 562 |        "      <td>399.0</td>\n",
 563 |        "      <td>0.0</td>\n",
 564 |        "    </tr>\n",
 565 |        "    <tr>\n",
 566 |        "      <th>LOCAL_CUR_TRANS_TX_NUM</th>\n",
 567 |        "      <td>17241.0</td>\n",
 568 |        "      <td>0.177890</td>\n",
 569 |        "      <td>4.125816</td>\n",
 570 |        "      <td>0.0</td>\n",
 571 |        "      <td>0.0</td>\n",
 572 |        "      <td>0.0</td>\n",
 573 |        "      <td>0.0</td>\n",
 574 |        "      <td>429.0</td>\n",
 575 |        "      <td>0.0</td>\n",
 576 |        "    </tr>\n",
 577 |        "    <tr>\n",
 578 |        "      <th>LOCAL_CUR_LASTSAV_TX_NUM</th>\n",
 579 |        "      <td>17241.0</td>\n",
 580 |        "      <td>0.425381</td>\n",
 581 |        "      <td>2.319546</td>\n",
 582 |        "      <td>0.0</td>\n",
 583 |        "      <td>0.0</td>\n",
 584 |        "      <td>0.0</td>\n",
 585 |        "      <td>0.0</td>\n",
 586 |        "      <td>112.0</td>\n",
 587 |        "      <td>0.0</td>\n",
 588 |        "    </tr>\n",
 589 |        "    <tr>\n",
 590 |        "      <th>LOCAL_CUR_WITHDRAW_TX_NUM</th>\n",
 591 |        "      <td>17241.0</td>\n",
 592 |        "      <td>0.823560</td>\n",
 593 |        "      <td>2.577986</td>\n",
 594 |        "      <td>0.0</td>\n",
 595 |        "      <td>0.0</td>\n",
 596 |        "      <td>0.0</td>\n",
 597 |        "      <td>1.0</td>\n",
 598 |        "      <td>138.0</td>\n",
 599 |        "      <td>0.0</td>\n",
 600 |        "    </tr>\n",
 601 |        "    <tr>\n",
 602 |        "      <th>LOCAL_FIX_OPEN_ACC_TX_NUM</th>\n",
 603 |        "      <td>17241.0</td>\n",
 604 |        "      <td>0.159910</td>\n",
 605 |        "      <td>0.470605</td>\n",
 606 |        "      <td>0.0</td>\n",
 607 |        "      <td>0.0</td>\n",
 608 |        "      <td>0.0</td>\n",
 609 |        "      <td>0.0</td>\n",
 610 |        "      <td>7.0</td>\n",
 611 |        "      <td>0.0</td>\n",
 612 |        "    </tr>\n",
 613 |        "    <tr>\n",
 614 |        "      <th>LOCAL_FIX_WITHDRAW_TX_NUM</th>\n",
 615 |        "      <td>17241.0</td>\n",
 616 |        "      <td>0.061539</td>\n",
 617 |        "      <td>0.496607</td>\n",
 618 |        "      <td>0.0</td>\n",
 619 |        "      <td>0.0</td>\n",
 620 |        "      <td>0.0</td>\n",
 621 |        "      <td>0.0</td>\n",
 622 |        "      <td>32.0</td>\n",
 623 |        "      <td>0.0</td>\n",
 624 |        "    </tr>\n",
 625 |        "    <tr>\n",
 626 |        "      <th>LOCAL_FIX_CLOSE_ACC_TX_NUM</th>\n",
 627 |        "      <td>17241.0</td>\n",
 628 |        "      <td>0.108578</td>\n",
 629 |        "      <td>0.452035</td>\n",
 630 |        "      <td>0.0</td>\n",
 631 |        "      <td>0.0</td>\n",
 632 |        "      <td>0.0</td>\n",
 633 |        "      <td>0.0</td>\n",
 634 |        "      <td>15.0</td>\n",
 635 |        "      <td>0.0</td>\n",
 636 |        "    </tr>\n",
 637 |        "    <tr>\n",
 638 |        "      <th>L6M_INDFINA_ALL_TX_NUM</th>\n",
 639 |        "      <td>17241.0</td>\n",
 640 |        "      <td>0.067571</td>\n",
 641 |        "      <td>0.923849</td>\n",
 642 |        "      <td>0.0</td>\n",
 643 |        "      <td>0.0</td>\n",
 644 |        "      <td>0.0</td>\n",
 645 |        "      <td>0.0</td>\n",
 646 |        "      <td>82.0</td>\n",
 647 |        "      <td>0.0</td>\n",
 648 |        "    </tr>\n",
 649 |        "    <tr>\n",
 650 |        "      <th>POS_CONSUME_TX_AMT</th>\n",
 651 |        "      <td>17241.0</td>\n",
 652 |        "      <td>1050.976915</td>\n",
 653 |        "      <td>23755.577944</td>\n",
 654 |        "      <td>0.0</td>\n",
 655 |        "      <td>0.0</td>\n",
 656 |        "      <td>0.0</td>\n",
 657 |        "      <td>0.0</td>\n",
 658 |        "      <td>1794863.0</td>\n",
 659 |        "      <td>0.0</td>\n",
 660 |        "    </tr>\n",
 661 |        "    <tr>\n",
 662 |        "      <th>POS_CONSUME_TX_NUM</th>\n",
 663 |        "      <td>17241.0</td>\n",
 664 |        "      <td>0.106896</td>\n",
 665 |        "      <td>0.819569</td>\n",
 666 |        "      <td>0.0</td>\n",
 667 |        "      <td>0.0</td>\n",
 668 |        "      <td>0.0</td>\n",
 669 |        "      <td>0.0</td>\n",
 670 |        "      <td>50.0</td>\n",
 671 |        "      <td>0.0</td>\n",
 672 |        "    </tr>\n",
 673 |        "    <tr>\n",
 674 |        "      <th>ATM_ACCT_TX_NUM</th>\n",
 675 |        "      <td>17241.0</td>\n",
 676 |        "      <td>0.409605</td>\n",
 677 |        "      <td>4.681921</td>\n",
 678 |        "      <td>0.0</td>\n",
 679 |        "      <td>0.0</td>\n",
 680 |        "      <td>0.0</td>\n",
 681 |        "      <td>0.0</td>\n",
 682 |        "      <td>257.0</td>\n",
 683 |        "      <td>0.0</td>\n",
 684 |        "    </tr>\n",
 685 |        "    <tr>\n",
 686 |        "      <th>ATM_ACCT_TX_AMT</th>\n",
 687 |        "      <td>17241.0</td>\n",
 688 |        "      <td>5632.666551</td>\n",
 689 |        "      <td>208094.147877</td>\n",
 690 |        "      <td>0.0</td>\n",
 691 |        "      <td>0.0</td>\n",
 692 |        "      <td>0.0</td>\n",
 693 |        "      <td>0.0</td>\n",
 694 |        "      <td>18501490.0</td>\n",
 695 |        "      <td>0.0</td>\n",
 696 |        "    </tr>\n",
 697 |        "    <tr>\n",
 698 |        "      <th>ATM_NOT_ACCT_TX_NUM</th>\n",
 699 |        "      <td>17241.0</td>\n",
 700 |        "      <td>0.188156</td>\n",
 701 |        "      <td>0.872419</td>\n",
 702 |        "      <td>0.0</td>\n",
 703 |        "      <td>0.0</td>\n",
 704 |        "      <td>0.0</td>\n",
 705 |        "      <td>0.0</td>\n",
 706 |        "      <td>31.0</td>\n",
 707 |        "      <td>0.0</td>\n",
 708 |        "    </tr>\n",
 709 |        "    <tr>\n",
 710 |        "      <th>ATM_ALL_TX_NUM</th>\n",
 711 |        "      <td>17241.0</td>\n",
 712 |        "      <td>0.597761</td>\n",
 713 |        "      <td>4.974549</td>\n",
 714 |        "      <td>0.0</td>\n",
 715 |        "      <td>0.0</td>\n",
 716 |        "      <td>0.0</td>\n",
 717 |        "      <td>0.0</td>\n",
 718 |        "      <td>261.0</td>\n",
 719 |        "      <td>0.0</td>\n",
 720 |        "    </tr>\n",
 721 |        "    <tr>\n",
 722 |        "      <th>COUNTER_NOT_ACCT_TX_NUM</th>\n",
 723 |        "      <td>17241.0</td>\n",
 724 |        "      <td>0.322313</td>\n",
 725 |        "      <td>0.870298</td>\n",
 726 |        "      <td>0.0</td>\n",
 727 |        "      <td>0.0</td>\n",
 728 |        "      <td>0.0</td>\n",
 729 |        "      <td>0.0</td>\n",
 730 |        "      <td>24.0</td>\n",
 731 |        "      <td>0.0</td>\n",
 732 |        "    </tr>\n",
 733 |        "    <tr>\n",
 734 |        "      <th>COUNTER_ACCT_TX_AMT</th>\n",
 735 |        "      <td>17241.0</td>\n",
 736 |        "      <td>32726.390987</td>\n",
 737 |        "      <td>153197.540483</td>\n",
 738 |        "      <td>0.0</td>\n",
 739 |        "      <td>0.0</td>\n",
 740 |        "      <td>0.0</td>\n",
 741 |        "      <td>12000.0</td>\n",
 742 |        "      <td>6402993.0</td>\n",
 743 |        "      <td>0.0</td>\n",
 744 |        "    </tr>\n",
 745 |        "    <tr>\n",
 746 |        "      <th>COUNTER_ACCT_TX_NUM</th>\n",
 747 |        "      <td>17241.0</td>\n",
 748 |        "      <td>1.303347</td>\n",
 749 |        "      <td>4.093381</td>\n",
 750 |        "      <td>0.0</td>\n",
 751 |        "      <td>0.0</td>\n",
 752 |        "      <td>0.0</td>\n",
 753 |        "      <td>2.0</td>\n",
 754 |        "      <td>307.0</td>\n",
 755 |        "      <td>0.0</td>\n",
 756 |        "    </tr>\n",
 757 |        "    <tr>\n",
 758 |        "      <th>COUNTER_ALL_TX_NUM</th>\n",
 759 |        "      <td>17241.0</td>\n",
 760 |        "      <td>1.625660</td>\n",
 761 |        "      <td>4.560893</td>\n",
 762 |        "      <td>0.0</td>\n",
 763 |        "      <td>0.0</td>\n",
 764 |        "      <td>0.0</td>\n",
 765 |        "      <td>2.0</td>\n",
 766 |        "      <td>331.0</td>\n",
 767 |        "      <td>0.0</td>\n",
 768 |        "    </tr>\n",
 769 |        "    <tr>\n",
 770 |        "      <th>NAT_DEBT_OPEN_ACC_DUR</th>\n",
 771 |        "      <td>17241.0</td>\n",
 772 |        "      <td>-1262.148367</td>\n",
 773 |        "      <td>10929.030591</td>\n",
 774 |        "      <td>-95877.0</td>\n",
 775 |        "      <td>0.0</td>\n",
 776 |        "      <td>0.0</td>\n",
 777 |        "      <td>0.0</td>\n",
 778 |        "      <td>85.0</td>\n",
 779 |        "      <td>0.0</td>\n",
 780 |        "    </tr>\n",
 781 |        "    <tr>\n",
 782 |        "      <th>FINA_OPEN_ACC_DUR</th>\n",
 783 |        "      <td>17241.0</td>\n",
 784 |        "      <td>-1017.557914</td>\n",
 785 |        "      <td>9825.499880</td>\n",
 786 |        "      <td>-95877.0</td>\n",
 787 |        "      <td>0.0</td>\n",
 788 |        "      <td>0.0</td>\n",
 789 |        "      <td>0.0</td>\n",
 790 |        "      <td>24.0</td>\n",
 791 |        "      <td>0.0</td>\n",
 792 |        "    </tr>\n",
 793 |        "    <tr>\n",
 794 |        "      <th>FUND_OPEN_ACC_DUR</th>\n",
 795 |        "      <td>17241.0</td>\n",
 796 |        "      <td>-967.307001</td>\n",
 797 |        "      <td>9583.391244</td>\n",
 798 |        "      <td>-95877.0</td>\n",
 799 |        "      <td>0.0</td>\n",
 800 |        "      <td>0.0</td>\n",
 801 |        "      <td>0.0</td>\n",
 802 |        "      <td>43.0</td>\n",
 803 |        "      <td>0.0</td>\n",
 804 |        "    </tr>\n",
 805 |        "    <tr>\n",
 806 |        "      <th>TELEBANK_ALL_TX_NUM</th>\n",
 807 |        "      <td>17241.0</td>\n",
 808 |        "      <td>0.000000</td>\n",
 809 |        "      <td>0.000000</td>\n",
 810 |        "      <td>0.0</td>\n",
 811 |        "      <td>0.0</td>\n",
 812 |        "      <td>0.0</td>\n",
 813 |        "      <td>0.0</td>\n",
 814 |        "      <td>0.0</td>\n",
 815 |        "      <td>0.0</td>\n",
 816 |        "    </tr>\n",
 817 |        "    <tr>\n",
 818 |        "      <th>CHURN_CUST_IND</th>\n",
 819 |        "      <td>17241.0</td>\n",
 820 |        "      <td>0.100980</td>\n",
 821 |        "      <td>0.301311</td>\n",
 822 |        "      <td>0.0</td>\n",
 823 |        "      <td>0.0</td>\n",
 824 |        "      <td>0.0</td>\n",
 825 |        "      <td>0.0</td>\n",
 826 |        "      <td>1.0</td>\n",
 827 |        "      <td>0.0</td>\n",
 828 |        "    </tr>\n",
 829 |        "    <tr>\n",
 830 |        "      <th>unan_vce_Range</th>\n",
 831 |        "      <td>17241.0</td>\n",
 832 |        "      <td>21.107650</td>\n",
 833 |        "      <td>35.917786</td>\n",
 834 |        "      <td>0.0</td>\n",
 835 |        "      <td>4.0</td>\n",
 836 |        "      <td>11.0</td>\n",
 837 |        "      <td>25.0</td>\n",
 838 |        "      <td>1395.0</td>\n",
 839 |        "      <td>0.0</td>\n",
 840 |        "    </tr>\n",
 841 |        "    <tr>\n",
 842 |        "      <th>unan_dat_Range</th>\n",
 843 |        "      <td>17241.0</td>\n",
 844 |        "      <td>0.064961</td>\n",
 845 |        "      <td>0.899576</td>\n",
 846 |        "      <td>0.0</td>\n",
 847 |        "      <td>0.0</td>\n",
 848 |        "      <td>0.0</td>\n",
 849 |        "      <td>0.0</td>\n",
 850 |        "      <td>83.0</td>\n",
 851 |        "      <td>0.0</td>\n",
 852 |        "    </tr>\n",
 853 |        "    <tr>\n",
 854 |        "      <th>callfwdv_Range</th>\n",
 855 |        "      <td>17241.0</td>\n",
 856 |        "      <td>0.025927</td>\n",
 857 |        "      <td>0.809077</td>\n",
 858 |        "      <td>0.0</td>\n",
 859 |        "      <td>0.0</td>\n",
 860 |        "      <td>0.0</td>\n",
 861 |        "      <td>0.0</td>\n",
 862 |        "      <td>59.0</td>\n",
 863 |        "      <td>0.0</td>\n",
 864 |        "    </tr>\n",
 865 |        "    <tr>\n",
 866 |        "      <th>totcalls</th>\n",
 867 |        "      <td>17241.0</td>\n",
 868 |        "      <td>2976.027377</td>\n",
 869 |        "      <td>4147.062069</td>\n",
 870 |        "      <td>0.0</td>\n",
 871 |        "      <td>853.0</td>\n",
 872 |        "      <td>1792.0</td>\n",
 873 |        "      <td>3564.0</td>\n",
 874 |        "      <td>98874.0</td>\n",
 875 |        "      <td>0.0</td>\n",
 876 |        "    </tr>\n",
 877 |        "    <tr>\n",
 878 |        "      <th>avg3mou</th>\n",
 879 |        "      <td>17241.0</td>\n",
 880 |        "      <td>546.483441</td>\n",
 881 |        "      <td>554.652467</td>\n",
 882 |        "      <td>0.0</td>\n",
 883 |        "      <td>166.0</td>\n",
 884 |        "      <td>376.0</td>\n",
 885 |        "      <td>746.0</td>\n",
 886 |        "      <td>5821.0</td>\n",
 887 |        "      <td>0.0</td>\n",
 888 |        "    </tr>\n",
 889 |        "    <tr>\n",
 890 |        "      <th>avg3qty</th>\n",
 891 |        "      <td>17241.0</td>\n",
 892 |        "      <td>190.337799</td>\n",
 893 |        "      <td>205.371296</td>\n",
 894 |        "      <td>0.0</td>\n",
 895 |        "      <td>59.0</td>\n",
 896 |        "      <td>131.0</td>\n",
 897 |        "      <td>249.0</td>\n",
 898 |        "      <td>3261.0</td>\n",
 899 |        "      <td>0.0</td>\n",
 900 |        "    </tr>\n",
 901 |        "    <tr>\n",
 902 |        "      <th>avg3rev</th>\n",
 903 |        "      <td>17241.0</td>\n",
 904 |        "      <td>60.086074</td>\n",
 905 |        "      <td>46.046675</td>\n",
 906 |        "      <td>1.0</td>\n",
 907 |        "      <td>34.0</td>\n",
 908 |        "      <td>49.0</td>\n",
 909 |        "      <td>72.0</td>\n",
 910 |        "      <td>835.0</td>\n",
 911 |        "      <td>0.0</td>\n",
 912 |        "    </tr>\n",
 913 |        "  </tbody>\n",
 914 |        "</table>\n",
 915 |        "</div>"
 916 |       ],
 917 |       "text/plain": [
 918 |        "                              count          mean            std      min  \\\n",
 919 |        "CUST_ID                     17241.0   8621.000000    4977.192331      1.0   \n",
 920 |        "OPEN_ACC_DUR                17241.0     80.104750      34.944320      0.0   \n",
 921 |        "AGE                         17241.0     49.339481      15.431282      4.0   \n",
 922 |        "LOCAL_CUR_ACCT_NUM          17241.0      1.796880       3.503054      0.0   \n",
 923 |        "LOCAL_CUR_TRANS_TX_NUM      17241.0      0.177890       4.125816      0.0   \n",
 924 |        "LOCAL_CUR_LASTSAV_TX_NUM    17241.0      0.425381       2.319546      0.0   \n",
 925 |        "LOCAL_CUR_WITHDRAW_TX_NUM   17241.0      0.823560       2.577986      0.0   \n",
 926 |        "LOCAL_FIX_OPEN_ACC_TX_NUM   17241.0      0.159910       0.470605      0.0   \n",
 927 |        "LOCAL_FIX_WITHDRAW_TX_NUM   17241.0      0.061539       0.496607      0.0   \n",
 928 |        "LOCAL_FIX_CLOSE_ACC_TX_NUM  17241.0      0.108578       0.452035      0.0   \n",
 929 |        "L6M_INDFINA_ALL_TX_NUM      17241.0      0.067571       0.923849      0.0   \n",
 930 |        "POS_CONSUME_TX_AMT          17241.0   1050.976915   23755.577944      0.0   \n",
 931 |        "POS_CONSUME_TX_NUM          17241.0      0.106896       0.819569      0.0   \n",
 932 |        "ATM_ACCT_TX_NUM             17241.0      0.409605       4.681921      0.0   \n",
 933 |        "ATM_ACCT_TX_AMT             17241.0   5632.666551  208094.147877      0.0   \n",
 934 |        "ATM_NOT_ACCT_TX_NUM         17241.0      0.188156       0.872419      0.0   \n",
 935 |        "ATM_ALL_TX_NUM              17241.0      0.597761       4.974549      0.0   \n",
 936 |        "COUNTER_NOT_ACCT_TX_NUM     17241.0      0.322313       0.870298      0.0   \n",
 937 |        "COUNTER_ACCT_TX_AMT         17241.0  32726.390987  153197.540483      0.0   \n",
 938 |        "COUNTER_ACCT_TX_NUM         17241.0      1.303347       4.093381      0.0   \n",
 939 |        "COUNTER_ALL_TX_NUM          17241.0      1.625660       4.560893      0.0   \n",
 940 |        "NAT_DEBT_OPEN_ACC_DUR       17241.0  -1262.148367   10929.030591 -95877.0   \n",
 941 |        "FINA_OPEN_ACC_DUR           17241.0  -1017.557914    9825.499880 -95877.0   \n",
 942 |        "FUND_OPEN_ACC_DUR           17241.0   -967.307001    9583.391244 -95877.0   \n",
 943 |        "TELEBANK_ALL_TX_NUM         17241.0      0.000000       0.000000      0.0   \n",
 944 |        "CHURN_CUST_IND              17241.0      0.100980       0.301311      0.0   \n",
 945 |        "unan_vce_Range              17241.0     21.107650      35.917786      0.0   \n",
 946 |        "unan_dat_Range              17241.0      0.064961       0.899576      0.0   \n",
 947 |        "callfwdv_Range              17241.0      0.025927       0.809077      0.0   \n",
 948 |        "totcalls                    17241.0   2976.027377    4147.062069      0.0   \n",
 949 |        "avg3mou                     17241.0    546.483441     554.652467      0.0   \n",
 950 |        "avg3qty                     17241.0    190.337799     205.371296      0.0   \n",
 951 |        "avg3rev                     17241.0     60.086074      46.046675      1.0   \n",
 952 |        "\n",
 953 |        "                               25%     50%      75%         max  missing_pct  \n",
 954 |        "CUST_ID                     4311.0  8621.0  12931.0     17241.0          0.0  \n",
 955 |        "OPEN_ACC_DUR                  56.0    75.0    104.0       278.0          0.0  \n",
 956 |        "AGE                           38.0    47.0     59.0       110.0          0.0  \n",
 957 |        "LOCAL_CUR_ACCT_NUM             1.0     1.0      2.0       399.0          0.0  \n",
 958 |        "LOCAL_CUR_TRANS_TX_NUM         0.0     0.0      0.0       429.0          0.0  \n",
 959 |        "LOCAL_CUR_LASTSAV_TX_NUM       0.0     0.0      0.0       112.0          0.0  \n",
 960 |        "LOCAL_CUR_WITHDRAW_TX_NUM      0.0     0.0      1.0       138.0          0.0  \n",
 961 |        "LOCAL_FIX_OPEN_ACC_TX_NUM      0.0     0.0      0.0         7.0          0.0  \n",
 962 |        "LOCAL_FIX_WITHDRAW_TX_NUM      0.0     0.0      0.0        32.0          0.0  \n",
 963 |        "LOCAL_FIX_CLOSE_ACC_TX_NUM     0.0     0.0      0.0        15.0          0.0  \n",
 964 |        "L6M_INDFINA_ALL_TX_NUM         0.0     0.0      0.0        82.0          0.0  \n",
 965 |        "POS_CONSUME_TX_AMT             0.0     0.0      0.0   1794863.0          0.0  \n",
 966 |        "POS_CONSUME_TX_NUM             0.0     0.0      0.0        50.0          0.0  \n",
 967 |        "ATM_ACCT_TX_NUM                0.0     0.0      0.0       257.0          0.0  \n",
 968 |        "ATM_ACCT_TX_AMT                0.0     0.0      0.0  18501490.0          0.0  \n",
 969 |        "ATM_NOT_ACCT_TX_NUM            0.0     0.0      0.0        31.0          0.0  \n",
 970 |        "ATM_ALL_TX_NUM                 0.0     0.0      0.0       261.0          0.0  \n",
 971 |        "COUNTER_NOT_ACCT_TX_NUM        0.0     0.0      0.0        24.0          0.0  \n",
 972 |        "COUNTER_ACCT_TX_AMT            0.0     0.0  12000.0   6402993.0          0.0  \n",
 973 |        "COUNTER_ACCT_TX_NUM            0.0     0.0      2.0       307.0          0.0  \n",
 974 |        "COUNTER_ALL_TX_NUM             0.0     0.0      2.0       331.0          0.0  \n",
 975 |        "NAT_DEBT_OPEN_ACC_DUR          0.0     0.0      0.0        85.0          0.0  \n",
 976 |        "FINA_OPEN_ACC_DUR              0.0     0.0      0.0        24.0          0.0  \n",
 977 |        "FUND_OPEN_ACC_DUR              0.0     0.0      0.0        43.0          0.0  \n",
 978 |        "TELEBANK_ALL_TX_NUM            0.0     0.0      0.0         0.0          0.0  \n",
 979 |        "CHURN_CUST_IND                 0.0     0.0      0.0         1.0          0.0  \n",
 980 |        "unan_vce_Range                 4.0    11.0     25.0      1395.0          0.0  \n",
 981 |        "unan_dat_Range                 0.0     0.0      0.0        83.0          0.0  \n",
 982 |        "callfwdv_Range                 0.0     0.0      0.0        59.0          0.0  \n",
 983 |        "totcalls                     853.0  1792.0   3564.0     98874.0          0.0  \n",
 984 |        "avg3mou                      166.0   376.0    746.0      5821.0          0.0  \n",
 985 |        "avg3qty                       59.0   131.0    249.0      3261.0          0.0  \n",
 986 |        "avg3rev                       34.0    49.0     72.0       835.0          0.0  "
 987 |       ]
 988 |      },
 989 |      "execution_count": 6,
 990 |      "metadata": {},
 991 |      "output_type": "execute_result"
 992 |     }
 993 |    ],
 994 |    "source": [
 995 |     "int_df = model_data.select_dtypes(include=['int64']).describe().T.assign(missing_pct=model_data.apply(lambda x: (len(x) - x.count())/float(len(x))))\n",
 996 |     "int_df"
 997 |    ]
 998 |   },
 999 |   {
1000 |    "cell_type": "code",
1001 |    "execution_count": 7,
1002 |    "metadata": {},
1003 |    "outputs": [],
1004 |    "source": [
1005 |     "del model_data['CUST_ID']"
1006 |    ]
1007 |   },
1008 |   {
1009 |    "cell_type": "code",
1010 |    "execution_count": 8,
1011 |    "metadata": {},
1012 |    "outputs": [
1013 |     {
1014 |      "data": {
1015 |       "text/html": [
1016 |        "<div>\n",
1017 |        "<style>\n",
1018 |        "    .dataframe thead tr:only-child th {\n",
1019 |        "        text-align: right;\n",
1020 |        "    }\n",
1021 |        "\n",
1022 |        "    .dataframe thead th {\n",
1023 |        "        text-align: left;\n",
1024 |        "    }\n",
1025 |        "\n",
1026 |        "    .dataframe tbody tr th {\n",
1027 |        "        vertical-align: top;\n",
1028 |        "    }\n",
1029 |        "</style>\n",
1030 |        "<table border=\"1\" class=\"dataframe\">\n",
1031 |        "  <thead>\n",
1032 |        "    <tr style=\"text-align: right;\">\n",
1033 |        "      <th></th>\n",
1034 |        "      <th>count</th>\n",
1035 |        "      <th>mean</th>\n",
1036 |        "      <th>std</th>\n",
1037 |        "      <th>min</th>\n",
1038 |        "      <th>25%</th>\n",
1039 |        "      <th>50%</th>\n",
1040 |        "      <th>75%</th>\n",
1041 |        "      <th>max</th>\n",
1042 |        "      <th>missing_pct</th>\n",
1043 |        "    </tr>\n",
1044 |        "  </thead>\n",
1045 |        "  <tbody>\n",
1046 |        "    <tr>\n",
1047 |        "      <th>LOCAL_CUR_SAV_SLOPE</th>\n",
1048 |        "      <td>17241.0</td>\n",
1049 |        "      <td>-0.011627</td>\n",
1050 |        "      <td>0.407344</td>\n",
1051 |        "      <td>-1.000000</td>\n",
1052 |        "      <td>-0.087996</td>\n",
1053 |        "      <td>0.000000</td>\n",
1054 |        "      <td>0.074506</td>\n",
1055 |        "      <td>2.000000e+00</td>\n",
1056 |        "      <td>0.000000</td>\n",
1057 |        "    </tr>\n",
1058 |        "    <tr>\n",
1059 |        "      <th>totmou</th>\n",
1060 |        "      <td>17241.0</td>\n",
1061 |        "      <td>7842.995152</td>\n",
1062 |        "      <td>9244.876680</td>\n",
1063 |        "      <td>0.000000</td>\n",
1064 |        "      <td>2450.000000</td>\n",
1065 |        "      <td>5098.000000</td>\n",
1066 |        "      <td>9868.000000</td>\n",
1067 |        "      <td>1.736084e+05</td>\n",
1068 |        "      <td>0.000000</td>\n",
1069 |        "    </tr>\n",
1070 |        "    <tr>\n",
1071 |        "      <th>mou_opkd_Mean</th>\n",
1072 |        "      <td>17241.0</td>\n",
1073 |        "      <td>1.230280</td>\n",
1074 |        "      <td>24.995118</td>\n",
1075 |        "      <td>0.000000</td>\n",
1076 |        "      <td>0.000000</td>\n",
1077 |        "      <td>0.000000</td>\n",
1078 |        "      <td>0.000000</td>\n",
1079 |        "      <td>2.922043e+03</td>\n",
1080 |        "      <td>0.000000</td>\n",
1081 |        "    </tr>\n",
1082 |        "    <tr>\n",
1083 |        "      <th>mou_opkv_Mean</th>\n",
1084 |        "      <td>17241.0</td>\n",
1085 |        "      <td>175.394446</td>\n",
1086 |        "      <td>243.886548</td>\n",
1087 |        "      <td>0.000000</td>\n",
1088 |        "      <td>21.876667</td>\n",
1089 |        "      <td>83.176667</td>\n",
1090 |        "      <td>229.656667</td>\n",
1091 |        "      <td>2.687313e+03</td>\n",
1092 |        "      <td>0.000000</td>\n",
1093 |        "    </tr>\n",
1094 |        "    <tr>\n",
1095 |        "      <th>opk_dat_Mean</th>\n",
1096 |        "      <td>17241.0</td>\n",
1097 |        "      <td>0.418402</td>\n",
1098 |        "      <td>4.264581</td>\n",
1099 |        "      <td>0.000000</td>\n",
1100 |        "      <td>0.000000</td>\n",
1101 |        "      <td>0.000000</td>\n",
1102 |        "      <td>0.000000</td>\n",
1103 |        "      <td>2.456667e+02</td>\n",
1104 |        "      <td>0.000000</td>\n",
1105 |        "    </tr>\n",
1106 |        "    <tr>\n",
1107 |        "      <th>opk_vce_Mean</th>\n",
1108 |        "      <td>17241.0</td>\n",
1109 |        "      <td>70.563444</td>\n",
1110 |        "      <td>98.128628</td>\n",
1111 |        "      <td>0.000000</td>\n",
1112 |        "      <td>11.666667</td>\n",
1113 |        "      <td>36.666667</td>\n",
1114 |        "      <td>91.666667</td>\n",
1115 |        "      <td>1.438000e+03</td>\n",
1116 |        "      <td>0.000000</td>\n",
1117 |        "    </tr>\n",
1118 |        "    <tr>\n",
1119 |        "      <th>mou_pead_Mean</th>\n",
1120 |        "      <td>17241.0</td>\n",
1121 |        "      <td>0.697439</td>\n",
1122 |        "      <td>6.788500</td>\n",
1123 |        "      <td>0.000000</td>\n",
1124 |        "      <td>0.000000</td>\n",
1125 |        "      <td>0.000000</td>\n",
1126 |        "      <td>0.000000</td>\n",
1127 |        "      <td>2.902433e+02</td>\n",
1128 |        "      <td>0.000000</td>\n",
1129 |        "    </tr>\n",
1130 |        "    <tr>\n",
1131 |        "      <th>mou_peav_Mean</th>\n",
1132 |        "      <td>17241.0</td>\n",
1133 |        "      <td>183.094184</td>\n",
1134 |        "      <td>213.011899</td>\n",
1135 |        "      <td>0.000000</td>\n",
1136 |        "      <td>43.940000</td>\n",
1137 |        "      <td>122.703333</td>\n",
1138 |        "      <td>242.296667</td>\n",
1139 |        "      <td>2.994580e+03</td>\n",
1140 |        "      <td>0.000000</td>\n",
1141 |        "    </tr>\n",
1142 |        "    <tr>\n",
1143 |        "      <th>peak_dat_Mean</th>\n",
1144 |        "      <td>17241.0</td>\n",
1145 |        "      <td>0.357829</td>\n",
1146 |        "      <td>3.878087</td>\n",
1147 |        "      <td>0.000000</td>\n",
1148 |        "      <td>0.000000</td>\n",
1149 |        "      <td>0.000000</td>\n",
1150 |        "      <td>0.000000</td>\n",
1151 |        "      <td>2.806667e+02</td>\n",
1152 |        "      <td>0.000000</td>\n",
1153 |        "    </tr>\n",
1154 |        "    <tr>\n",
1155 |        "      <th>peak_vce_Mean</th>\n",
1156 |        "      <td>17241.0</td>\n",
1157 |        "      <td>93.384316</td>\n",
1158 |        "      <td>107.881729</td>\n",
1159 |        "      <td>0.000000</td>\n",
1160 |        "      <td>24.333333</td>\n",
1161 |        "      <td>63.666667</td>\n",
1162 |        "      <td>124.333333</td>\n",
1163 |        "      <td>1.921333e+03</td>\n",
1164 |        "      <td>0.000000</td>\n",
1165 |        "    </tr>\n",
1166 |        "    <tr>\n",
1167 |        "      <th>mouiwylisv_Mean</th>\n",
1168 |        "      <td>17241.0</td>\n",
1169 |        "      <td>19.708691</td>\n",
1170 |        "      <td>42.675059</td>\n",
1171 |        "      <td>0.000000</td>\n",
1172 |        "      <td>0.000000</td>\n",
1173 |        "      <td>3.983333</td>\n",
1174 |        "      <td>20.666667</td>\n",
1175 |        "      <td>9.210700e+02</td>\n",
1176 |        "      <td>0.000000</td>\n",
1177 |        "    </tr>\n",
1178 |        "    <tr>\n",
1179 |        "      <th>iwylis_vce_Mean</th>\n",
1180 |        "      <td>17241.0</td>\n",
1181 |        "      <td>8.615722</td>\n",
1182 |        "      <td>17.234011</td>\n",
1183 |        "      <td>0.000000</td>\n",
1184 |        "      <td>0.000000</td>\n",
1185 |        "      <td>2.333333</td>\n",
1186 |        "      <td>9.666667</td>\n",
1187 |        "      <td>3.446667e+02</td>\n",
1188 |        "      <td>0.000000</td>\n",
1189 |        "    </tr>\n",
1190 |        "    <tr>\n",
1191 |        "      <th>mouowylisv_Mean</th>\n",
1192 |        "      <td>17241.0</td>\n",
1193 |        "      <td>30.144799</td>\n",
1194 |        "      <td>50.859667</td>\n",
1195 |        "      <td>0.000000</td>\n",
1196 |        "      <td>2.860000</td>\n",
1197 |        "      <td>13.016667</td>\n",
1198 |        "      <td>36.880000</td>\n",
1199 |        "      <td>1.490253e+03</td>\n",
1200 |        "      <td>0.000000</td>\n",
1201 |        "    </tr>\n",
1202 |        "    <tr>\n",
1203 |        "      <th>owylis_vce_Mean</th>\n",
1204 |        "      <td>17241.0</td>\n",
1205 |        "      <td>26.186648</td>\n",
1206 |        "      <td>35.982014</td>\n",
1207 |        "      <td>0.000000</td>\n",
1208 |        "      <td>3.666667</td>\n",
1209 |        "      <td>14.000000</td>\n",
1210 |        "      <td>35.000000</td>\n",
1211 |        "      <td>6.443333e+02</td>\n",
1212 |        "      <td>0.000000</td>\n",
1213 |        "    </tr>\n",
1214 |        "    <tr>\n",
1215 |        "      <th>mou_rvce_Mean</th>\n",
1216 |        "      <td>17241.0</td>\n",
1217 |        "      <td>119.739875</td>\n",
1218 |        "      <td>170.938653</td>\n",
1219 |        "      <td>0.000000</td>\n",
1220 |        "      <td>9.466667</td>\n",
1221 |        "      <td>54.290000</td>\n",
1222 |        "      <td>161.336667</td>\n",
1223 |        "      <td>2.138510e+03</td>\n",
1224 |        "      <td>0.000000</td>\n",
1225 |        "    </tr>\n",
1226 |        "    <tr>\n",
1227 |        "      <th>mou_cdat_Mean</th>\n",
1228 |        "      <td>17241.0</td>\n",
1229 |        "      <td>1.930040</td>\n",
1230 |        "      <td>28.330685</td>\n",
1231 |        "      <td>0.000000</td>\n",
1232 |        "      <td>0.000000</td>\n",
1233 |        "      <td>0.000000</td>\n",
1234 |        "      <td>0.000000</td>\n",
1235 |        "      <td>3.032050e+03</td>\n",
1236 |        "      <td>0.000000</td>\n",
1237 |        "    </tr>\n",
1238 |        "    <tr>\n",
1239 |        "      <th>unan_vce_Mean</th>\n",
1240 |        "      <td>17241.0</td>\n",
1241 |        "      <td>28.921389</td>\n",
1242 |        "      <td>39.442498</td>\n",
1243 |        "      <td>0.000000</td>\n",
1244 |        "      <td>5.333333</td>\n",
1245 |        "      <td>16.666667</td>\n",
1246 |        "      <td>37.000000</td>\n",
1247 |        "      <td>8.143333e+02</td>\n",
1248 |        "      <td>0.000000</td>\n",
1249 |        "    </tr>\n",
1250 |        "    <tr>\n",
1251 |        "      <th>unan_dat_Mean</th>\n",
1252 |        "      <td>17241.0</td>\n",
1253 |        "      <td>0.031417</td>\n",
1254 |        "      <td>0.458737</td>\n",
1255 |        "      <td>0.000000</td>\n",
1256 |        "      <td>0.000000</td>\n",
1257 |        "      <td>0.000000</td>\n",
1258 |        "      <td>0.000000</td>\n",
1259 |        "      <td>4.133333e+01</td>\n",
1260 |        "      <td>0.000000</td>\n",
1261 |        "    </tr>\n",
1262 |        "    <tr>\n",
1263 |        "      <th>comp_vce_Mean</th>\n",
1264 |        "      <td>17241.0</td>\n",
1265 |        "      <td>114.545889</td>\n",
1266 |        "      <td>122.214639</td>\n",
1267 |        "      <td>0.000000</td>\n",
1268 |        "      <td>31.666667</td>\n",
1269 |        "      <td>79.666667</td>\n",
1270 |        "      <td>157.333333</td>\n",
1271 |        "      <td>1.376667e+03</td>\n",
1272 |        "      <td>0.000000</td>\n",
1273 |        "    </tr>\n",
1274 |        "    <tr>\n",
1275 |        "      <th>comp_dat_Mean</th>\n",
1276 |        "      <td>17241.0</td>\n",
1277 |        "      <td>0.776231</td>\n",
1278 |        "      <td>7.644959</td>\n",
1279 |        "      <td>0.000000</td>\n",
1280 |        "      <td>0.000000</td>\n",
1281 |        "      <td>0.000000</td>\n",
1282 |        "      <td>0.000000</td>\n",
1283 |        "      <td>5.263333e+02</td>\n",
1284 |        "      <td>0.000000</td>\n",
1285 |        "    </tr>\n",
1286 |        "    <tr>\n",
1287 |        "      <th>custcare_Mean</th>\n",
1288 |        "      <td>17241.0</td>\n",
1289 |        "      <td>1.949056</td>\n",
1290 |        "      <td>5.958278</td>\n",
1291 |        "      <td>0.000000</td>\n",
1292 |        "      <td>0.000000</td>\n",
1293 |        "      <td>0.000000</td>\n",
1294 |        "      <td>1.666667</td>\n",
1295 |        "      <td>3.656667e+02</td>\n",
1296 |        "      <td>0.000000</td>\n",
1297 |        "    </tr>\n",
1298 |        "    <tr>\n",
1299 |        "      <th>cc_mou_Mean</th>\n",
1300 |        "      <td>17241.0</td>\n",
1301 |        "      <td>3.995940</td>\n",
1302 |        "      <td>11.376720</td>\n",
1303 |        "      <td>0.000000</td>\n",
1304 |        "      <td>0.000000</td>\n",
1305 |        "      <td>0.000000</td>\n",
1306 |        "      <td>3.190000</td>\n",
1307 |        "      <td>3.091133e+02</td>\n",
1308 |        "      <td>0.000000</td>\n",
1309 |        "    </tr>\n",
1310 |        "    <tr>\n",
1311 |        "      <th>avgmou</th>\n",
1312 |        "      <td>17241.0</td>\n",
1313 |        "      <td>494.788836</td>\n",
1314 |        "      <td>452.298412</td>\n",
1315 |        "      <td>0.000000</td>\n",
1316 |        "      <td>179.710000</td>\n",
1317 |        "      <td>365.000000</td>\n",
1318 |        "      <td>667.330000</td>\n",
1319 |        "      <td>5.164290e+03</td>\n",
1320 |        "      <td>0.000000</td>\n",
1321 |        "    </tr>\n",
1322 |        "    <tr>\n",
1323 |        "      <th>avgqty</th>\n",
1324 |        "      <td>17241.0</td>\n",
1325 |        "      <td>179.168816</td>\n",
1326 |        "      <td>179.200014</td>\n",
1327 |        "      <td>0.000000</td>\n",
1328 |        "      <td>64.300000</td>\n",
1329 |        "      <td>129.170000</td>\n",
1330 |        "      <td>234.680000</td>\n",
1331 |        "      <td>2.500890e+03</td>\n",
1332 |        "      <td>0.000000</td>\n",
1333 |        "    </tr>\n",
1334 |        "    <tr>\n",
1335 |        "      <th>L6M_INDFINA_ALL_TX_AMT</th>\n",
1336 |        "      <td>17241.0</td>\n",
1337 |        "      <td>4303.363895</td>\n",
1338 |        "      <td>76935.023181</td>\n",
1339 |        "      <td>0.000000</td>\n",
1340 |        "      <td>0.000000</td>\n",
1341 |        "      <td>0.000000</td>\n",
1342 |        "      <td>0.000000</td>\n",
1343 |        "      <td>3.987167e+06</td>\n",
1344 |        "      <td>0.000000</td>\n",
1345 |        "    </tr>\n",
1346 |        "    <tr>\n",
1347 |        "      <th>LOCAL_FIX_CLOSE_ACC_TX_AMT</th>\n",
1348 |        "      <td>17241.0</td>\n",
1349 |        "      <td>3453.773522</td>\n",
1350 |        "      <td>20665.105788</td>\n",
1351 |        "      <td>0.000000</td>\n",
1352 |        "      <td>0.000000</td>\n",
1353 |        "      <td>0.000000</td>\n",
1354 |        "      <td>0.000000</td>\n",
1355 |        "      <td>7.945000e+05</td>\n",
1356 |        "      <td>0.000000</td>\n",
1357 |        "    </tr>\n",
1358 |        "    <tr>\n",
1359 |        "      <th>LOCAL_CUR_MON_AVG_BAL</th>\n",
1360 |        "      <td>17241.0</td>\n",
1361 |        "      <td>27735.417899</td>\n",
1362 |        "      <td>98608.195518</td>\n",
1363 |        "      <td>0.000000</td>\n",
1364 |        "      <td>51.530000</td>\n",
1365 |        "      <td>3232.020000</td>\n",
1366 |        "      <td>23113.460000</td>\n",
1367 |        "      <td>7.610110e+06</td>\n",
1368 |        "      <td>0.000000</td>\n",
1369 |        "    </tr>\n",
1370 |        "    <tr>\n",
1371 |        "      <th>LOCAL_CUR_MON_AVG_BAL_PROP</th>\n",
1372 |        "      <td>17241.0</td>\n",
1373 |        "      <td>0.373216</td>\n",
1374 |        "      <td>0.438130</td>\n",
1375 |        "      <td>0.000000</td>\n",
1376 |        "      <td>0.005796</td>\n",
1377 |        "      <td>0.091914</td>\n",
1378 |        "      <td>1.000000</td>\n",
1379 |        "      <td>1.000000e+00</td>\n",
1380 |        "      <td>0.000000</td>\n",
1381 |        "    </tr>\n",
1382 |        "    <tr>\n",
1383 |        "      <th>LOCAL_OVEONEYR_FF_MON_AVG_BAL</th>\n",
1384 |        "      <td>17241.0</td>\n",
1385 |        "      <td>72023.920103</td>\n",
1386 |        "      <td>128981.784041</td>\n",
1387 |        "      <td>0.000000</td>\n",
1388 |        "      <td>0.000000</td>\n",
1389 |        "      <td>50000.000000</td>\n",
1390 |        "      <td>95000.000000</td>\n",
1391 |        "      <td>5.561334e+06</td>\n",
1392 |        "      <td>0.000000</td>\n",
1393 |        "    </tr>\n",
1394 |        "    <tr>\n",
1395 |        "      <th>LOCAL_FIX_MON_AVG_BAL</th>\n",
1396 |        "      <td>17241.0</td>\n",
1397 |        "      <td>83907.649074</td>\n",
1398 |        "      <td>360269.646860</td>\n",
1399 |        "      <td>0.000000</td>\n",
1400 |        "      <td>0.000000</td>\n",
1401 |        "      <td>56677.270000</td>\n",
1402 |        "      <td>101125.000000</td>\n",
1403 |        "      <td>4.358232e+07</td>\n",
1404 |        "      <td>0.000000</td>\n",
1405 |        "    </tr>\n",
1406 |        "    <tr>\n",
1407 |        "      <th>...</th>\n",
1408 |        "      <td>...</td>\n",
1409 |        "      <td>...</td>\n",
1410 |        "      <td>...</td>\n",
1411 |        "      <td>...</td>\n",
1412 |        "      <td>...</td>\n",
1413 |        "      <td>...</td>\n",
1414 |        "      <td>...</td>\n",
1415 |        "      <td>...</td>\n",
1416 |        "      <td>...</td>\n",
1417 |        "    </tr>\n",
1418 |        "    <tr>\n",
1419 |        "      <th>LOCAL_FIX_OPEN_ACC_TX_AMT</th>\n",
1420 |        "      <td>17241.0</td>\n",
1421 |        "      <td>5528.391594</td>\n",
1422 |        "      <td>33922.675509</td>\n",
1423 |        "      <td>0.000000</td>\n",
1424 |        "      <td>0.000000</td>\n",
1425 |        "      <td>0.000000</td>\n",
1426 |        "      <td>0.000000</td>\n",
1427 |        "      <td>2.586000e+06</td>\n",
1428 |        "      <td>0.000000</td>\n",
1429 |        "    </tr>\n",
1430 |        "    <tr>\n",
1431 |        "      <th>LOCAL_FIX_WITHDRAW_TX_AMT</th>\n",
1432 |        "      <td>17241.0</td>\n",
1433 |        "      <td>1706.323595</td>\n",
1434 |        "      <td>18679.511773</td>\n",
1435 |        "      <td>0.000000</td>\n",
1436 |        "      <td>0.000000</td>\n",
1437 |        "      <td>0.000000</td>\n",
1438 |        "      <td>0.000000</td>\n",
1439 |        "      <td>1.398000e+06</td>\n",
1440 |        "      <td>0.000000</td>\n",
1441 |        "    </tr>\n",
1442 |        "    <tr>\n",
1443 |        "      <th>SAV_SLOPE</th>\n",
1444 |        "      <td>17241.0</td>\n",
1445 |        "      <td>-0.004043</td>\n",
1446 |        "      <td>0.256375</td>\n",
1447 |        "      <td>-1.000000</td>\n",
1448 |        "      <td>-0.000550</td>\n",
1449 |        "      <td>0.000533</td>\n",
1450 |        "      <td>0.035337</td>\n",
1451 |        "      <td>1.904013e+00</td>\n",
1452 |        "      <td>0.000000</td>\n",
1453 |        "    </tr>\n",
1454 |        "    <tr>\n",
1455 |        "      <th>mou_cvce_Mean</th>\n",
1456 |        "      <td>17241.0</td>\n",
1457 |        "      <td>238.802099</td>\n",
1458 |        "      <td>267.274943</td>\n",
1459 |        "      <td>0.000000</td>\n",
1460 |        "      <td>56.900000</td>\n",
1461 |        "      <td>157.183333</td>\n",
1462 |        "      <td>329.523333</td>\n",
1463 |        "      <td>3.661677e+03</td>\n",
1464 |        "      <td>0.000000</td>\n",
1465 |        "    </tr>\n",
1466 |        "    <tr>\n",
1467 |        "      <th>vceovr_Range</th>\n",
1468 |        "      <td>17205.0</td>\n",
1469 |        "      <td>29.374902</td>\n",
1470 |        "      <td>58.098543</td>\n",
1471 |        "      <td>0.000000</td>\n",
1472 |        "      <td>0.000000</td>\n",
1473 |        "      <td>2.100000</td>\n",
1474 |        "      <td>37.000000</td>\n",
1475 |        "      <td>1.215550e+03</td>\n",
1476 |        "      <td>0.002088</td>\n",
1477 |        "    </tr>\n",
1478 |        "    <tr>\n",
1479 |        "      <th>vceovr_Mean</th>\n",
1480 |        "      <td>17205.0</td>\n",
1481 |        "      <td>12.848716</td>\n",
1482 |        "      <td>29.276031</td>\n",
1483 |        "      <td>0.000000</td>\n",
1484 |        "      <td>0.000000</td>\n",
1485 |        "      <td>0.600000</td>\n",
1486 |        "      <td>13.275000</td>\n",
1487 |        "      <td>6.012125e+02</td>\n",
1488 |        "      <td>0.002088</td>\n",
1489 |        "    </tr>\n",
1490 |        "    <tr>\n",
1491 |        "      <th>roam_Range</th>\n",
1492 |        "      <td>17205.0</td>\n",
1493 |        "      <td>3.127544</td>\n",
1494 |        "      <td>19.692447</td>\n",
1495 |        "      <td>0.000000</td>\n",
1496 |        "      <td>0.000000</td>\n",
1497 |        "      <td>0.000000</td>\n",
1498 |        "      <td>0.780000</td>\n",
1499 |        "      <td>1.561470e+03</td>\n",
1500 |        "      <td>0.002088</td>\n",
1501 |        "    </tr>\n",
1502 |        "    <tr>\n",
1503 |        "      <th>datovr_Range</th>\n",
1504 |        "      <td>17205.0</td>\n",
1505 |        "      <td>0.717182</td>\n",
1506 |        "      <td>5.964260</td>\n",
1507 |        "      <td>0.000000</td>\n",
1508 |        "      <td>0.000000</td>\n",
1509 |        "      <td>0.000000</td>\n",
1510 |        "      <td>0.000000</td>\n",
1511 |        "      <td>3.032000e+02</td>\n",
1512 |        "      <td>0.002088</td>\n",
1513 |        "    </tr>\n",
1514 |        "    <tr>\n",
1515 |        "      <th>da_Mean</th>\n",
1516 |        "      <td>17205.0</td>\n",
1517 |        "      <td>0.920998</td>\n",
1518 |        "      <td>2.118237</td>\n",
1519 |        "      <td>0.000000</td>\n",
1520 |        "      <td>0.000000</td>\n",
1521 |        "      <td>0.247500</td>\n",
1522 |        "      <td>0.990000</td>\n",
1523 |        "      <td>5.766750e+01</td>\n",
1524 |        "      <td>0.002088</td>\n",
1525 |        "    </tr>\n",
1526 |        "    <tr>\n",
1527 |        "      <th>ovrrev_Range</th>\n",
1528 |        "      <td>17205.0</td>\n",
1529 |        "      <td>29.844381</td>\n",
1530 |        "      <td>58.365586</td>\n",
1531 |        "      <td>0.000000</td>\n",
1532 |        "      <td>0.000000</td>\n",
1533 |        "      <td>3.300000</td>\n",
1534 |        "      <td>37.450000</td>\n",
1535 |        "      <td>1.215550e+03</td>\n",
1536 |        "      <td>0.002088</td>\n",
1537 |        "    </tr>\n",
1538 |        "    <tr>\n",
1539 |        "      <th>ovrmou_Range</th>\n",
1540 |        "      <td>17205.0</td>\n",
1541 |        "      <td>89.809416</td>\n",
1542 |        "      <td>177.656676</td>\n",
1543 |        "      <td>0.000000</td>\n",
1544 |        "      <td>0.000000</td>\n",
1545 |        "      <td>9.000000</td>\n",
1546 |        "      <td>112.000000</td>\n",
1547 |        "      <td>3.473000e+03</td>\n",
1548 |        "      <td>0.002088</td>\n",
1549 |        "    </tr>\n",
1550 |        "    <tr>\n",
1551 |        "      <th>da_Range</th>\n",
1552 |        "      <td>17205.0</td>\n",
1553 |        "      <td>1.671578</td>\n",
1554 |        "      <td>2.987940</td>\n",
1555 |        "      <td>0.000000</td>\n",
1556 |        "      <td>0.000000</td>\n",
1557 |        "      <td>0.990000</td>\n",
1558 |        "      <td>1.980000</td>\n",
1559 |        "      <td>6.732000e+01</td>\n",
1560 |        "      <td>0.002088</td>\n",
1561 |        "    </tr>\n",
1562 |        "    <tr>\n",
1563 |        "      <th>totmrc_Range</th>\n",
1564 |        "      <td>17205.0</td>\n",
1565 |        "      <td>8.801175</td>\n",
1566 |        "      <td>27.175065</td>\n",
1567 |        "      <td>0.000000</td>\n",
1568 |        "      <td>0.000000</td>\n",
1569 |        "      <td>0.000000</td>\n",
1570 |        "      <td>0.000000</td>\n",
1571 |        "      <td>5.999800e+02</td>\n",
1572 |        "      <td>0.002088</td>\n",
1573 |        "    </tr>\n",
1574 |        "    <tr>\n",
1575 |        "      <th>mou_Range</th>\n",
1576 |        "      <td>17205.0</td>\n",
1577 |        "      <td>369.560622</td>\n",
1578 |        "      <td>420.042160</td>\n",
1579 |        "      <td>0.000000</td>\n",
1580 |        "      <td>116.000000</td>\n",
1581 |        "      <td>242.000000</td>\n",
1582 |        "      <td>468.000000</td>\n",
1583 |        "      <td>6.865000e+03</td>\n",
1584 |        "      <td>0.002088</td>\n",
1585 |        "    </tr>\n",
1586 |        "    <tr>\n",
1587 |        "      <th>rev_Range</th>\n",
1588 |        "      <td>17205.0</td>\n",
1589 |        "      <td>42.799289</td>\n",
1590 |        "      <td>70.035631</td>\n",
1591 |        "      <td>0.000000</td>\n",
1592 |        "      <td>1.980000</td>\n",
1593 |        "      <td>15.750000</td>\n",
1594 |        "      <td>55.570000</td>\n",
1595 |        "      <td>1.527970e+03</td>\n",
1596 |        "      <td>0.002088</td>\n",
1597 |        "    </tr>\n",
1598 |        "    <tr>\n",
1599 |        "      <th>roam_Mean</th>\n",
1600 |        "      <td>17205.0</td>\n",
1601 |        "      <td>1.142233</td>\n",
1602 |        "      <td>6.544958</td>\n",
1603 |        "      <td>0.000000</td>\n",
1604 |        "      <td>0.000000</td>\n",
1605 |        "      <td>0.000000</td>\n",
1606 |        "      <td>0.257500</td>\n",
1607 |        "      <td>4.162575e+02</td>\n",
1608 |        "      <td>0.002088</td>\n",
1609 |        "    </tr>\n",
1610 |        "    <tr>\n",
1611 |        "      <th>datovr_Mean</th>\n",
1612 |        "      <td>17205.0</td>\n",
1613 |        "      <td>0.249300</td>\n",
1614 |        "      <td>2.132651</td>\n",
1615 |        "      <td>0.000000</td>\n",
1616 |        "      <td>0.000000</td>\n",
1617 |        "      <td>0.000000</td>\n",
1618 |        "      <td>0.000000</td>\n",
1619 |        "      <td>1.007000e+02</td>\n",
1620 |        "      <td>0.002088</td>\n",
1621 |        "    </tr>\n",
1622 |        "    <tr>\n",
1623 |        "      <th>ovrmou_Mean</th>\n",
1624 |        "      <td>17205.0</td>\n",
1625 |        "      <td>39.815027</td>\n",
1626 |        "      <td>92.879658</td>\n",
1627 |        "      <td>0.000000</td>\n",
1628 |        "      <td>0.000000</td>\n",
1629 |        "      <td>2.750000</td>\n",
1630 |        "      <td>41.000000</td>\n",
1631 |        "      <td>1.887250e+03</td>\n",
1632 |        "      <td>0.002088</td>\n",
1633 |        "    </tr>\n",
1634 |        "    <tr>\n",
1635 |        "      <th>ovrrev_Mean</th>\n",
1636 |        "      <td>17205.0</td>\n",
1637 |        "      <td>13.101876</td>\n",
1638 |        "      <td>29.493960</td>\n",
1639 |        "      <td>0.000000</td>\n",
1640 |        "      <td>0.000000</td>\n",
1641 |        "      <td>0.975000</td>\n",
1642 |        "      <td>13.700000</td>\n",
1643 |        "      <td>6.012125e+02</td>\n",
1644 |        "      <td>0.002088</td>\n",
1645 |        "    </tr>\n",
1646 |        "    <tr>\n",
1647 |        "      <th>change_mou</th>\n",
1648 |        "      <td>17153.0</td>\n",
1649 |        "      <td>-8.395409</td>\n",
1650 |        "      <td>248.516433</td>\n",
1651 |        "      <td>-3875.000000</td>\n",
1652 |        "      <td>-78.250000</td>\n",
1653 |        "      <td>-3.500000</td>\n",
1654 |        "      <td>70.000000</td>\n",
1655 |        "      <td>3.712250e+03</td>\n",
1656 |        "      <td>0.005104</td>\n",
1657 |        "    </tr>\n",
1658 |        "    <tr>\n",
1659 |        "      <th>change_rev</th>\n",
1660 |        "      <td>17153.0</td>\n",
1661 |        "      <td>-1.693052</td>\n",
1662 |        "      <td>34.284273</td>\n",
1663 |        "      <td>-626.357500</td>\n",
1664 |        "      <td>-7.082500</td>\n",
1665 |        "      <td>-0.312500</td>\n",
1666 |        "      <td>1.612500</td>\n",
1667 |        "      <td>4.191725e+02</td>\n",
1668 |        "      <td>0.005104</td>\n",
1669 |        "    </tr>\n",
1670 |        "    <tr>\n",
1671 |        "      <th>hnd_price</th>\n",
1672 |        "      <td>17043.0</td>\n",
1673 |        "      <td>106.805692</td>\n",
1674 |        "      <td>61.622379</td>\n",
1675 |        "      <td>9.989998</td>\n",
1676 |        "      <td>59.989990</td>\n",
1677 |        "      <td>129.989990</td>\n",
1678 |        "      <td>149.989990</td>\n",
1679 |        "      <td>4.999900e+02</td>\n",
1680 |        "      <td>0.011484</td>\n",
1681 |        "    </tr>\n",
1682 |        "    <tr>\n",
1683 |        "      <th>forgntvl</th>\n",
1684 |        "      <td>16935.0</td>\n",
1685 |        "      <td>0.058636</td>\n",
1686 |        "      <td>0.234949</td>\n",
1687 |        "      <td>0.000000</td>\n",
1688 |        "      <td>0.000000</td>\n",
1689 |        "      <td>0.000000</td>\n",
1690 |        "      <td>0.000000</td>\n",
1691 |        "      <td>1.000000e+00</td>\n",
1692 |        "      <td>0.017748</td>\n",
1693 |        "    </tr>\n",
1694 |        "    <tr>\n",
1695 |        "      <th>age1</th>\n",
1696 |        "      <td>16935.0</td>\n",
1697 |        "      <td>31.669324</td>\n",
1698 |        "      <td>22.129690</td>\n",
1699 |        "      <td>0.000000</td>\n",
1700 |        "      <td>0.000000</td>\n",
1701 |        "      <td>36.000000</td>\n",
1702 |        "      <td>48.000000</td>\n",
1703 |        "      <td>9.600000e+01</td>\n",
1704 |        "      <td>0.017748</td>\n",
1705 |        "    </tr>\n",
1706 |        "    <tr>\n",
1707 |        "      <th>age2</th>\n",
1708 |        "      <td>16935.0</td>\n",
1709 |        "      <td>21.225037</td>\n",
1710 |        "      <td>23.903794</td>\n",
1711 |        "      <td>0.000000</td>\n",
1712 |        "      <td>0.000000</td>\n",
1713 |        "      <td>0.000000</td>\n",
1714 |        "      <td>42.000000</td>\n",
1715 |        "      <td>9.800000e+01</td>\n",
1716 |        "      <td>0.017748</td>\n",
1717 |        "    </tr>\n",
1718 |        "    <tr>\n",
1719 |        "      <th>avg6qty</th>\n",
1720 |        "      <td>16688.0</td>\n",
1721 |        "      <td>187.114334</td>\n",
1722 |        "      <td>195.036224</td>\n",
1723 |        "      <td>0.000000</td>\n",
1724 |        "      <td>61.000000</td>\n",
1725 |        "      <td>131.000000</td>\n",
1726 |        "      <td>246.000000</td>\n",
1727 |        "      <td>2.673000e+03</td>\n",
1728 |        "      <td>0.032075</td>\n",
1729 |        "    </tr>\n",
1730 |        "    <tr>\n",
1731 |        "      <th>avg6mou</th>\n",
1732 |        "      <td>16688.0</td>\n",
1733 |        "      <td>531.920662</td>\n",
1734 |        "      <td>517.294774</td>\n",
1735 |        "      <td>0.000000</td>\n",
1736 |        "      <td>173.000000</td>\n",
1737 |        "      <td>374.000000</td>\n",
1738 |        "      <td>726.250000</td>\n",
1739 |        "      <td>5.347000e+03</td>\n",
1740 |        "      <td>0.032075</td>\n",
1741 |        "    </tr>\n",
1742 |        "    <tr>\n",
1743 |        "      <th>lor</th>\n",
1744 |        "      <td>12230.0</td>\n",
1745 |        "      <td>6.377514</td>\n",
1746 |        "      <td>4.781650</td>\n",
1747 |        "      <td>0.000000</td>\n",
1748 |        "      <td>2.000000</td>\n",
1749 |        "      <td>5.000000</td>\n",
1750 |        "      <td>10.000000</td>\n",
1751 |        "      <td>1.500000e+01</td>\n",
1752 |        "      <td>0.290644</td>\n",
1753 |        "    </tr>\n",
1754 |        "    <tr>\n",
1755 |        "      <th>pre_hnd_price</th>\n",
1756 |        "      <td>7649.0</td>\n",
1757 |        "      <td>82.761595</td>\n",
1758 |        "      <td>60.803160</td>\n",
1759 |        "      <td>9.989998</td>\n",
1760 |        "      <td>29.989990</td>\n",
1761 |        "      <td>59.989990</td>\n",
1762 |        "      <td>129.989990</td>\n",
1763 |        "      <td>4.999900e+02</td>\n",
1764 |        "      <td>0.556348</td>\n",
1765 |        "    </tr>\n",
1766 |        "    <tr>\n",
1767 |        "      <th>educ1</th>\n",
1768 |        "      <td>2419.0</td>\n",
1769 |        "      <td>1.868541</td>\n",
1770 |        "      <td>0.847854</td>\n",
1771 |        "      <td>1.000000</td>\n",
1772 |        "      <td>1.000000</td>\n",
1773 |        "      <td>2.000000</td>\n",
1774 |        "      <td>2.000000</td>\n",
1775 |        "      <td>4.000000e+00</td>\n",
1776 |        "      <td>0.859695</td>\n",
1777 |        "    </tr>\n",
1778 |        "  </tbody>\n",
1779 |        "</table>\n",
1780 |        "<p>76 rows × 9 columns</p>\n",
1781 |        "</div>"
1782 |       ],
1783 |       "text/plain": [
1784 |        "                                 count          mean            std  \\\n",
1785 |        "LOCAL_CUR_SAV_SLOPE            17241.0     -0.011627       0.407344   \n",
1786 |        "totmou                         17241.0   7842.995152    9244.876680   \n",
1787 |        "mou_opkd_Mean                  17241.0      1.230280      24.995118   \n",
1788 |        "mou_opkv_Mean                  17241.0    175.394446     243.886548   \n",
1789 |        "opk_dat_Mean                   17241.0      0.418402       4.264581   \n",
1790 |        "opk_vce_Mean                   17241.0     70.563444      98.128628   \n",
1791 |        "mou_pead_Mean                  17241.0      0.697439       6.788500   \n",
1792 |        "mou_peav_Mean                  17241.0    183.094184     213.011899   \n",
1793 |        "peak_dat_Mean                  17241.0      0.357829       3.878087   \n",
1794 |        "peak_vce_Mean                  17241.0     93.384316     107.881729   \n",
1795 |        "mouiwylisv_Mean                17241.0     19.708691      42.675059   \n",
1796 |        "iwylis_vce_Mean                17241.0      8.615722      17.234011   \n",
1797 |        "mouowylisv_Mean                17241.0     30.144799      50.859667   \n",
1798 |        "owylis_vce_Mean                17241.0     26.186648      35.982014   \n",
1799 |        "mou_rvce_Mean                  17241.0    119.739875     170.938653   \n",
1800 |        "mou_cdat_Mean                  17241.0      1.930040      28.330685   \n",
1801 |        "unan_vce_Mean                  17241.0     28.921389      39.442498   \n",
1802 |        "unan_dat_Mean                  17241.0      0.031417       0.458737   \n",
1803 |        "comp_vce_Mean                  17241.0    114.545889     122.214639   \n",
1804 |        "comp_dat_Mean                  17241.0      0.776231       7.644959   \n",
1805 |        "custcare_Mean                  17241.0      1.949056       5.958278   \n",
1806 |        "cc_mou_Mean                    17241.0      3.995940      11.376720   \n",
1807 |        "avgmou                         17241.0    494.788836     452.298412   \n",
1808 |        "avgqty                         17241.0    179.168816     179.200014   \n",
1809 |        "L6M_INDFINA_ALL_TX_AMT         17241.0   4303.363895   76935.023181   \n",
1810 |        "LOCAL_FIX_CLOSE_ACC_TX_AMT     17241.0   3453.773522   20665.105788   \n",
1811 |        "LOCAL_CUR_MON_AVG_BAL          17241.0  27735.417899   98608.195518   \n",
1812 |        "LOCAL_CUR_MON_AVG_BAL_PROP     17241.0      0.373216       0.438130   \n",
1813 |        "LOCAL_OVEONEYR_FF_MON_AVG_BAL  17241.0  72023.920103  128981.784041   \n",
1814 |        "LOCAL_FIX_MON_AVG_BAL          17241.0  83907.649074  360269.646860   \n",
1815 |        "...                                ...           ...            ...   \n",
1816 |        "LOCAL_FIX_OPEN_ACC_TX_AMT      17241.0   5528.391594   33922.675509   \n",
1817 |        "LOCAL_FIX_WITHDRAW_TX_AMT      17241.0   1706.323595   18679.511773   \n",
1818 |        "SAV_SLOPE                      17241.0     -0.004043       0.256375   \n",
1819 |        "mou_cvce_Mean                  17241.0    238.802099     267.274943   \n",
1820 |        "vceovr_Range                   17205.0     29.374902      58.098543   \n",
1821 |        "vceovr_Mean                    17205.0     12.848716      29.276031   \n",
1822 |        "roam_Range                     17205.0      3.127544      19.692447   \n",
1823 |        "datovr_Range                   17205.0      0.717182       5.964260   \n",
1824 |        "da_Mean                        17205.0      0.920998       2.118237   \n",
1825 |        "ovrrev_Range                   17205.0     29.844381      58.365586   \n",
1826 |        "ovrmou_Range                   17205.0     89.809416     177.656676   \n",
1827 |        "da_Range                       17205.0      1.671578       2.987940   \n",
1828 |        "totmrc_Range                   17205.0      8.801175      27.175065   \n",
1829 |        "mou_Range                      17205.0    369.560622     420.042160   \n",
1830 |        "rev_Range                      17205.0     42.799289      70.035631   \n",
1831 |        "roam_Mean                      17205.0      1.142233       6.544958   \n",
1832 |        "datovr_Mean                    17205.0      0.249300       2.132651   \n",
1833 |        "ovrmou_Mean                    17205.0     39.815027      92.879658   \n",
1834 |        "ovrrev_Mean                    17205.0     13.101876      29.493960   \n",
1835 |        "change_mou                     17153.0     -8.395409     248.516433   \n",
1836 |        "change_rev                     17153.0     -1.693052      34.284273   \n",
1837 |        "hnd_price                      17043.0    106.805692      61.622379   \n",
1838 |        "forgntvl                       16935.0      0.058636       0.234949   \n",
1839 |        "age1                           16935.0     31.669324      22.129690   \n",
1840 |        "age2                           16935.0     21.225037      23.903794   \n",
1841 |        "avg6qty                        16688.0    187.114334     195.036224   \n",
1842 |        "avg6mou                        16688.0    531.920662     517.294774   \n",
1843 |        "lor                            12230.0      6.377514       4.781650   \n",
1844 |        "pre_hnd_price                   7649.0     82.761595      60.803160   \n",
1845 |        "educ1                           2419.0      1.868541       0.847854   \n",
1846 |        "\n",
1847 |        "                                       min          25%           50%  \\\n",
1848 |        "LOCAL_CUR_SAV_SLOPE              -1.000000    -0.087996      0.000000   \n",
1849 |        "totmou                            0.000000  2450.000000   5098.000000   \n",
1850 |        "mou_opkd_Mean                     0.000000     0.000000      0.000000   \n",
1851 |        "mou_opkv_Mean                     0.000000    21.876667     83.176667   \n",
1852 |        "opk_dat_Mean                      0.000000     0.000000      0.000000   \n",
1853 |        "opk_vce_Mean                      0.000000    11.666667     36.666667   \n",
1854 |        "mou_pead_Mean                     0.000000     0.000000      0.000000   \n",
1855 |        "mou_peav_Mean                     0.000000    43.940000    122.703333   \n",
1856 |        "peak_dat_Mean                     0.000000     0.000000      0.000000   \n",
1857 |        "peak_vce_Mean                     0.000000    24.333333     63.666667   \n",
1858 |        "mouiwylisv_Mean                   0.000000     0.000000      3.983333   \n",
1859 |        "iwylis_vce_Mean                   0.000000     0.000000      2.333333   \n",
1860 |        "mouowylisv_Mean                   0.000000     2.860000     13.016667   \n",
1861 |        "owylis_vce_Mean                   0.000000     3.666667     14.000000   \n",
1862 |        "mou_rvce_Mean                     0.000000     9.466667     54.290000   \n",
1863 |        "mou_cdat_Mean                     0.000000     0.000000      0.000000   \n",
1864 |        "unan_vce_Mean                     0.000000     5.333333     16.666667   \n",
1865 |        "unan_dat_Mean                     0.000000     0.000000      0.000000   \n",
1866 |        "comp_vce_Mean                     0.000000    31.666667     79.666667   \n",
1867 |        "comp_dat_Mean                     0.000000     0.000000      0.000000   \n",
1868 |        "custcare_Mean                     0.000000     0.000000      0.000000   \n",
1869 |        "cc_mou_Mean                       0.000000     0.000000      0.000000   \n",
1870 |        "avgmou                            0.000000   179.710000    365.000000   \n",
1871 |        "avgqty                            0.000000    64.300000    129.170000   \n",
1872 |        "L6M_INDFINA_ALL_TX_AMT            0.000000     0.000000      0.000000   \n",
1873 |        "LOCAL_FIX_CLOSE_ACC_TX_AMT        0.000000     0.000000      0.000000   \n",
1874 |        "LOCAL_CUR_MON_AVG_BAL             0.000000    51.530000   3232.020000   \n",
1875 |        "LOCAL_CUR_MON_AVG_BAL_PROP        0.000000     0.005796      0.091914   \n",
1876 |        "LOCAL_OVEONEYR_FF_MON_AVG_BAL     0.000000     0.000000  50000.000000   \n",
1877 |        "LOCAL_FIX_MON_AVG_BAL             0.000000     0.000000  56677.270000   \n",
1878 |        "...                                    ...          ...           ...   \n",
1879 |        "LOCAL_FIX_OPEN_ACC_TX_AMT         0.000000     0.000000      0.000000   \n",
1880 |        "LOCAL_FIX_WITHDRAW_TX_AMT         0.000000     0.000000      0.000000   \n",
1881 |        "SAV_SLOPE                        -1.000000    -0.000550      0.000533   \n",
1882 |        "mou_cvce_Mean                     0.000000    56.900000    157.183333   \n",
1883 |        "vceovr_Range                      0.000000     0.000000      2.100000   \n",
1884 |        "vceovr_Mean                       0.000000     0.000000      0.600000   \n",
1885 |        "roam_Range                        0.000000     0.000000      0.000000   \n",
1886 |        "datovr_Range                      0.000000     0.000000      0.000000   \n",
1887 |        "da_Mean                           0.000000     0.000000      0.247500   \n",
1888 |        "ovrrev_Range                      0.000000     0.000000      3.300000   \n",
1889 |        "ovrmou_Range                      0.000000     0.000000      9.000000   \n",
1890 |        "da_Range                          0.000000     0.000000      0.990000   \n",
1891 |        "totmrc_Range                      0.000000     0.000000      0.000000   \n",
1892 |        "mou_Range                         0.000000   116.000000    242.000000   \n",
1893 |        "rev_Range                         0.000000     1.980000     15.750000   \n",
1894 |        "roam_Mean                         0.000000     0.000000      0.000000   \n",
1895 |        "datovr_Mean                       0.000000     0.000000      0.000000   \n",
1896 |        "ovrmou_Mean                       0.000000     0.000000      2.750000   \n",
1897 |        "ovrrev_Mean                       0.000000     0.000000      0.975000   \n",
1898 |        "change_mou                    -3875.000000   -78.250000     -3.500000   \n",
1899 |        "change_rev                     -626.357500    -7.082500     -0.312500   \n",
1900 |        "hnd_price                         9.989998    59.989990    129.989990   \n",
1901 |        "forgntvl                          0.000000     0.000000      0.000000   \n",
1902 |        "age1                              0.000000     0.000000     36.000000   \n",
1903 |        "age2                              0.000000     0.000000      0.000000   \n",
1904 |        "avg6qty                           0.000000    61.000000    131.000000   \n",
1905 |        "avg6mou                           0.000000   173.000000    374.000000   \n",
1906 |        "lor                               0.000000     2.000000      5.000000   \n",
1907 |        "pre_hnd_price                     9.989998    29.989990     59.989990   \n",
1908 |        "educ1                             1.000000     1.000000      2.000000   \n",
1909 |        "\n",
1910 |        "                                         75%           max  missing_pct  \n",
1911 |        "LOCAL_CUR_SAV_SLOPE                 0.074506  2.000000e+00     0.000000  \n",
1912 |        "totmou                           9868.000000  1.736084e+05     0.000000  \n",
1913 |        "mou_opkd_Mean                       0.000000  2.922043e+03     0.000000  \n",
1914 |        "mou_opkv_Mean                     229.656667  2.687313e+03     0.000000  \n",
1915 |        "opk_dat_Mean                        0.000000  2.456667e+02     0.000000  \n",
1916 |        "opk_vce_Mean                       91.666667  1.438000e+03     0.000000  \n",
1917 |        "mou_pead_Mean                       0.000000  2.902433e+02     0.000000  \n",
1918 |        "mou_peav_Mean                     242.296667  2.994580e+03     0.000000  \n",
1919 |        "peak_dat_Mean                       0.000000  2.806667e+02     0.000000  \n",
1920 |        "peak_vce_Mean                     124.333333  1.921333e+03     0.000000  \n",
1921 |        "mouiwylisv_Mean                    20.666667  9.210700e+02     0.000000  \n",
1922 |        "iwylis_vce_Mean                     9.666667  3.446667e+02     0.000000  \n",
1923 |        "mouowylisv_Mean                    36.880000  1.490253e+03     0.000000  \n",
1924 |        "owylis_vce_Mean                    35.000000  6.443333e+02     0.000000  \n",
1925 |        "mou_rvce_Mean                     161.336667  2.138510e+03     0.000000  \n",
1926 |        "mou_cdat_Mean                       0.000000  3.032050e+03     0.000000  \n",
1927 |        "unan_vce_Mean                      37.000000  8.143333e+02     0.000000  \n",
1928 |        "unan_dat_Mean                       0.000000  4.133333e+01     0.000000  \n",
1929 |        "comp_vce_Mean                     157.333333  1.376667e+03     0.000000  \n",
1930 |        "comp_dat_Mean                       0.000000  5.263333e+02     0.000000  \n",
1931 |        "custcare_Mean                       1.666667  3.656667e+02     0.000000  \n",
1932 |        "cc_mou_Mean                         3.190000  3.091133e+02     0.000000  \n",
1933 |        "avgmou                            667.330000  5.164290e+03     0.000000  \n",
1934 |        "avgqty                            234.680000  2.500890e+03     0.000000  \n",
1935 |        "L6M_INDFINA_ALL_TX_AMT              0.000000  3.987167e+06     0.000000  \n",
1936 |        "LOCAL_FIX_CLOSE_ACC_TX_AMT          0.000000  7.945000e+05     0.000000  \n",
1937 |        "LOCAL_CUR_MON_AVG_BAL           23113.460000  7.610110e+06     0.000000  \n",
1938 |        "LOCAL_CUR_MON_AVG_BAL_PROP          1.000000  1.000000e+00     0.000000  \n",
1939 |        "LOCAL_OVEONEYR_FF_MON_AVG_BAL   95000.000000  5.561334e+06     0.000000  \n",
1940 |        "LOCAL_FIX_MON_AVG_BAL          101125.000000  4.358232e+07     0.000000  \n",
1941 |        "...                                      ...           ...          ...  \n",
1942 |        "LOCAL_FIX_OPEN_ACC_TX_AMT           0.000000  2.586000e+06     0.000000  \n",
1943 |        "LOCAL_FIX_WITHDRAW_TX_AMT           0.000000  1.398000e+06     0.000000  \n",
1944 |        "SAV_SLOPE                           0.035337  1.904013e+00     0.000000  \n",
1945 |        "mou_cvce_Mean                     329.523333  3.661677e+03     0.000000  \n",
1946 |        "vceovr_Range                       37.000000  1.215550e+03     0.002088  \n",
1947 |        "vceovr_Mean                        13.275000  6.012125e+02     0.002088  \n",
1948 |        "roam_Range                          0.780000  1.561470e+03     0.002088  \n",
1949 |        "datovr_Range                        0.000000  3.032000e+02     0.002088  \n",
1950 |        "da_Mean                             0.990000  5.766750e+01     0.002088  \n",
1951 |        "ovrrev_Range                       37.450000  1.215550e+03     0.002088  \n",
1952 |        "ovrmou_Range                      112.000000  3.473000e+03     0.002088  \n",
1953 |        "da_Range                            1.980000  6.732000e+01     0.002088  \n",
1954 |        "totmrc_Range                        0.000000  5.999800e+02     0.002088  \n",
1955 |        "mou_Range                         468.000000  6.865000e+03     0.002088  \n",
1956 |        "rev_Range                          55.570000  1.527970e+03     0.002088  \n",
1957 |        "roam_Mean                           0.257500  4.162575e+02     0.002088  \n",
1958 |        "datovr_Mean                         0.000000  1.007000e+02     0.002088  \n",
1959 |        "ovrmou_Mean                        41.000000  1.887250e+03     0.002088  \n",
1960 |        "ovrrev_Mean                        13.700000  6.012125e+02     0.002088  \n",
1961 |        "change_mou                         70.000000  3.712250e+03     0.005104  \n",
1962 |        "change_rev                          1.612500  4.191725e+02     0.005104  \n",
1963 |        "hnd_price                         149.989990  4.999900e+02     0.011484  \n",
1964 |        "forgntvl                            0.000000  1.000000e+00     0.017748  \n",
1965 |        "age1                               48.000000  9.600000e+01     0.017748  \n",
1966 |        "age2                               42.000000  9.800000e+01     0.017748  \n",
1967 |        "avg6qty                           246.000000  2.673000e+03     0.032075  \n",
1968 |        "avg6mou                           726.250000  5.347000e+03     0.032075  \n",
1969 |        "lor                                10.000000  1.500000e+01     0.290644  \n",
1970 |        "pre_hnd_price                     129.989990  4.999900e+02     0.556348  \n",
1971 |        "educ1                               2.000000  4.000000e+00     0.859695  \n",
1972 |        "\n",
1973 |        "[76 rows x 9 columns]"
1974 |       ]
1975 |      },
1976 |      "execution_count": 8,
1977 |      "metadata": {},
1978 |      "output_type": "execute_result"
1979 |     }
1980 |    ],
1981 |    "source": [
1982 |     "float_df = model_data.select_dtypes(include=['float']).describe().T.assign(missing_pct=model_data.apply(lambda x: (len(x)-x.count())/float(len(x))))\n",
1983 |     "float_df = float_df.sort_values(['missing_pct'])\n",
1984 |     "float_df"
1985 |    ]
1986 |   },
1987 |   {
1988 |    "cell_type": "code",
1989 |    "execution_count": 9,
1990 |    "metadata": {},
1991 |    "outputs": [],
1992 |    "source": [
1993 |     "# 特征构造\n",
1994 |     "model_data['AVG_LOCAL_CUR_TRANS_TX_AMT'] = model_data['LOCAL_CUR_TRANS_TX_AMT'] / model_data['LOCAL_CUR_TRANS_TX_NUM']\n",
1995 |     "model_data['AVG_LOCAL_CUR_LASTSAV_TX_AMT'] = model_data['LOCAL_CUR_LASTSAV_TX_AMT'] / model_data['LOCAL_CUR_LASTSAV_TX_NUM']"
1996 |    ]
1997 |   },
1998 |   {
1999 |    "cell_type": "code",
2000 |    "execution_count": 10,
2001 |    "metadata": {},
2002 |    "outputs": [],
2003 |    "source": [
2004 |     "maxValueFeatures = ['LOCAL_CUR_SAV_SLOPE','LOCAL_BELONEYR_FF_SLOPE','LOCAL_OVEONEYR_FF_SLOPE','LOCAL_SAV_SLOPE','SAV_SLOPE']\n",
2005 |     "model_data['volatilityMax']= model_data[maxValueFeatures].apply(max, axis =1)"
2006 |    ]
2007 |   },
2008 |   {
2009 |    "cell_type": "code",
2010 |    "execution_count": 11,
2011 |    "metadata": {},
2012 |    "outputs": [],
2013 |    "source": [
2014 |     "# 本币活期月日均余额占比 = 1 - 本币定期月日均余额占比\n",
2015 |     "# 删除冗余特征\n",
2016 |     "del model_data['LOCAL_CUR_MON_AVG_BAL_PROP']"
2017 |    ]
2018 |   },
2019 |   {
2020 |    "cell_type": "code",
2021 |    "execution_count": 12,
2022 |    "metadata": {
2023 |     "collapsed": true
2024 |    },
2025 |    "outputs": [],
2026 |    "source": [
2027 |     "# 某些特征可以相加\n",
2028 |     "sumup_cols0 = ['LOCAL_CUR_MON_AVG_BAL','LOCAL_FIX_MON_AVG_BAL']\n",
2029 |     "sumup_cols1 = ['LOCAL_CUR_WITHDRAW_TX_NUM','LOCAL_FIX_WITHDRAW_TX_NUM']\n",
2030 |     "sumup_cols2 = ['LOCAL_CUR_WITHDRAW_TX_AMT','LOCAL_FIX_WITHDRAW_TX_AMT']\n",
2031 |     "sumup_cols3 = ['COUNTER_NOT_ACCT_TX_NUM','COUNTER_ACCT_TX_NUM']\n",
2032 |     "sumup_cols4 = ['ATM_ALL_TX_NUM','COUNTER_ALL_TX_NUM']\n",
2033 |     "sumup_cols5 = ['ATM_ACCT_TX_NUM','COUNTER_ACCT_TX_NUM']\n",
2034 |     "sumup_cols6 = ['ATM_ACCT_TX_AMT','COUNTER_ACCT_TX_AMT']\n",
2035 |     "sumup_cols7 = ['ATM_NOT_ACCT_TX_NUM','COUNTER_NOT_ACCT_TX_NUM']\n",
2036 |     "\n",
2037 |     "model_data['TOTAL_LOCAL_MON_AVG_BAL'] = model_data[sumup_cols0].apply(sum, axis = 1)\n",
2038 |     "model_data['TOTAL_WITHDRAW_TX_NUM'] = model_data[sumup_cols1].apply(sum, axis = 1)\n",
2039 |     "model_data['TOTAL_WITHDRAW_TX_AMT'] = model_data[sumup_cols2].apply(sum, axis = 1)\n",
2040 |     "model_data['TOTAL_COUNTER_TX_NUM'] = model_data[sumup_cols3].apply(sum, axis = 1)\n",
2041 |     "model_data['TOTAL_ALL_TX_NUM'] = model_data[sumup_cols4].apply(sum, axis = 1)\n",
2042 |     "model_data['TOTAL_ACCT_TX_NUM'] = model_data[sumup_cols5].apply(sum, axis = 1)\n",
2043 |     "model_data['TOTAL_ACCT_TX_AMT'] = model_data[sumup_cols6].apply(sum, axis = 1)\n",
2044 |     "model_data['TOTAL_NOT_ACCT_TX_NUM'] = model_data[sumup_cols7].apply(sum, axis = 1)"
2045 |    ]
2046 |   },
2047 |   {
2048 |    "cell_type": "code",
2049 |    "execution_count": 13,
2050 |    "metadata": {},
2051 |    "outputs": [
2052 |     {
2053 |      "data": {
2054 |       "text/plain": [
2055 |        "(17241, 133)"
2056 |       ]
2057 |      },
2058 |      "execution_count": 13,
2059 |      "metadata": {},
2060 |      "output_type": "execute_result"
2061 |     }
2062 |    ],
2063 |    "source": [
2064 |     "model_data.shape"
2065 |    ]
2066 |   },
2067 |   {
2068 |    "cell_type": "code",
2069 |    "execution_count": 14,
2070 |    "metadata": {
2071 |     "collapsed": true
2072 |    },
2073 |    "outputs": [],
2074 |    "source": [
2075 |     "### 特征构造: 比率\n",
2076 |     "numerator_cols = ['LOCAL_SAV_CUR_ALL_BAL','SAV_CUR_ALL_BAL','ASSET_CUR_ALL_BAL','LOCAL_CUR_WITHDRAW_TX_NUM','LOCAL_CUR_WITHDRAW_TX_AMT','COUNTER_NOT_ACCT_TX_NUM',\n",
2077 |     "                 'ATM_ALL_TX_NUM','ATM_ACCT_TX_AMT','ATM_NOT_ACCT_TX_NUM']\n",
2078 |     "denominator_cols = ['LOCAL_SAV_MON_AVG_BAL','SAV_MON_AVG_BAL','ASSET_MON_AVG_BAL','TOTAL_WITHDRAW_TX_NUM','TOTAL_WITHDRAW_TX_AMT','TOTAL_COUNTER_TX_NUM',\n",
2079 |     "                   'TOTAL_ACCT_TX_NUM','TOTAL_ACCT_TX_AMT','TOTAL_NOT_ACCT_TX_NUM']\n",
2080 |     "\n",
2081 |     "new_col_name = [\"RATIO_\"+str(i) for i in range(len(numerator_cols))]\n",
2082 |     "for i in range(len(numerator_cols)):\n",
2083 |     "    model_data[new_col_name[i]] = model_data[numerator_cols[i]] / model_data[denominator_cols[i]]"
2084 |    ]
2085 |   },
2086 |   {
2087 |    "cell_type": "code",
2088 |    "execution_count": 15,
2089 |    "metadata": {},
2090 |    "outputs": [
2091 |     {
2092 |      "data": {
2093 |       "text/plain": [
2094 |        "(17241, 142)"
2095 |       ]
2096 |      },
2097 |      "execution_count": 15,
2098 |      "metadata": {},
2099 |      "output_type": "execute_result"
2100 |     }
2101 |    ],
2102 |    "source": [
2103 |     "model_data.shape"
2104 |    ]
2105 |   },
2106 |   {
2107 |    "cell_type": "code",
2108 |    "execution_count": 16,
2109 |    "metadata": {
2110 |     "collapsed": true
2111 |    },
2112 |    "outputs": [],
2113 |    "source": [
2114 |     "object_df = model_data.select_dtypes(include=['object']).describe().T\n",
2115 |     "int_df = model_data.select_dtypes(include=['int64']).describe().T\n",
2116 |     "float_df = model_data.select_dtypes(include=['float']).describe().T"
2117 |    ]
2118 |   },
2119 |   {
2120 |    "cell_type": "code",
2121 |    "execution_count": 17,
2122 |    "metadata": {},
2123 |    "outputs": [],
2124 |    "source": [
2125 |     "# 对float和int类型的特征中的缺失值不全为0\n",
2126 |     "model_data_to_fillna = pd.concat([model_data[float_df.index], model_data[int_df.index]], axis=1)\n",
2127 |     "model_data_to_fillna.replace(float('inf'), 1, inplace=True)\n",
2128 |     "model_data_to_fillna.fillna(0, inplace=True)"
2129 |    ]
2130 |   },
2131 |   {
2132 |    "cell_type": "code",
2133 |    "execution_count": 18,
2134 |    "metadata": {},
2135 |    "outputs": [],
2136 |    "source": [
2137 |     "model_data = pd.concat([model_data[object_df.index], model_data_to_fillna], axis=1)"
2138 |    ]
2139 |   },
2140 |   {
2141 |    "cell_type": "code",
2142 |    "execution_count": 19,
2143 |    "metadata": {},
2144 |    "outputs": [],
2145 |    "source": [
2146 |     "# 进行One_hot编码，并且对object类型的特征中的缺失值都设定一个哑变量\n",
2147 |     "model_data = pd.get_dummies(model_data, dummy_na=True)"
2148 |    ]
2149 |   },
2150 |   {
2151 |    "cell_type": "code",
2152 |    "execution_count": 20,
2153 |    "metadata": {},
2154 |    "outputs": [
2155 |     {
2156 |      "data": {
2157 |       "text/html": [
2158 |        "<div>\n",
2159 |        "<style>\n",
2160 |        "    .dataframe thead tr:only-child th {\n",
2161 |        "        text-align: right;\n",
2162 |        "    }\n",
2163 |        "\n",
2164 |        "    .dataframe thead th {\n",
2165 |        "        text-align: left;\n",
2166 |        "    }\n",
2167 |        "\n",
2168 |        "    .dataframe tbody tr th {\n",
2169 |        "        vertical-align: top;\n",
2170 |        "    }\n",
2171 |        "</style>\n",
2172 |        "<table border=\"1\" class=\"dataframe\">\n",
2173 |        "  <thead>\n",
2174 |        "    <tr style=\"text-align: right;\">\n",
2175 |        "      <th></th>\n",
2176 |        "      <th>LOCAL_CUR_SAV_SLOPE</th>\n",
2177 |        "      <th>LOCAL_CUR_MON_AVG_BAL</th>\n",
2178 |        "      <th>LOCAL_OVEONEYR_FF_MON_AVG_BAL</th>\n",
2179 |        "      <th>LOCAL_FIX_MON_AVG_BAL</th>\n",
2180 |        "      <th>LOCAL_FIX_MON_AVG_BAL_PROP</th>\n",
2181 |        "      <th>LOCAL_BELONEYR_FF_SLOPE</th>\n",
2182 |        "      <th>LOCAL_BELONEYR_FF_MON_AVG_BAL</th>\n",
2183 |        "      <th>LOCAL_OVEONEYR_FF_SLOPE</th>\n",
2184 |        "      <th>LOCAL_SAV_SLOPE</th>\n",
2185 |        "      <th>LOCAL_SAV_CUR_ALL_BAL</th>\n",
2186 |        "      <th>...</th>\n",
2187 |        "      <th>kid6_10_nan</th>\n",
2188 |        "      <th>kid11_15_U</th>\n",
2189 |        "      <th>kid11_15_Y</th>\n",
2190 |        "      <th>kid11_15_nan</th>\n",
2191 |        "      <th>kid16_17_U</th>\n",
2192 |        "      <th>kid16_17_Y</th>\n",
2193 |        "      <th>kid16_17_nan</th>\n",
2194 |        "      <th>car_buy_New</th>\n",
2195 |        "      <th>car_buy_UNKNOWN</th>\n",
2196 |        "      <th>car_buy_nan</th>\n",
2197 |        "    </tr>\n",
2198 |        "  </thead>\n",
2199 |        "  <tbody>\n",
2200 |        "    <tr>\n",
2201 |        "      <th>0</th>\n",
2202 |        "      <td>0.000000</td>\n",
2203 |        "      <td>0.00</td>\n",
2204 |        "      <td>50429.68</td>\n",
2205 |        "      <td>50429.68</td>\n",
2206 |        "      <td>1.000000</td>\n",
2207 |        "      <td>0.0</td>\n",
2208 |        "      <td>0.0</td>\n",
2209 |        "      <td>-0.032395</td>\n",
2210 |        "      <td>-0.032395</td>\n",
2211 |        "      <td>41000.00</td>\n",
2212 |        "      <td>...</td>\n",
2213 |        "      <td>0</td>\n",
2214 |        "      <td>1</td>\n",
2215 |        "      <td>0</td>\n",
2216 |        "      <td>0</td>\n",
2217 |        "      <td>1</td>\n",
2218 |        "      <td>0</td>\n",
2219 |        "      <td>0</td>\n",
2220 |        "      <td>1</td>\n",
2221 |        "      <td>0</td>\n",
2222 |        "      <td>0</td>\n",
2223 |        "    </tr>\n",
2224 |        "    <tr>\n",
2225 |        "      <th>1</th>\n",
2226 |        "      <td>1.005692</td>\n",
2227 |        "      <td>20149.04</td>\n",
2228 |        "      <td>56047.50</td>\n",
2229 |        "      <td>56047.50</td>\n",
2230 |        "      <td>0.735565</td>\n",
2231 |        "      <td>0.0</td>\n",
2232 |        "      <td>0.0</td>\n",
2233 |        "      <td>-0.148857</td>\n",
2234 |        "      <td>0.003965</td>\n",
2235 |        "      <td>31929.10</td>\n",
2236 |        "      <td>...</td>\n",
2237 |        "      <td>0</td>\n",
2238 |        "      <td>1</td>\n",
2239 |        "      <td>0</td>\n",
2240 |        "      <td>0</td>\n",
2241 |        "      <td>1</td>\n",
2242 |        "      <td>0</td>\n",
2243 |        "      <td>0</td>\n",
2244 |        "      <td>1</td>\n",
2245 |        "      <td>0</td>\n",
2246 |        "      <td>0</td>\n",
2247 |        "    </tr>\n",
2248 |        "    <tr>\n",
2249 |        "      <th>2</th>\n",
2250 |        "      <td>0.000562</td>\n",
2251 |        "      <td>17.81</td>\n",
2252 |        "      <td>0.00</td>\n",
2253 |        "      <td>0.00</td>\n",
2254 |        "      <td>0.000000</td>\n",
2255 |        "      <td>0.0</td>\n",
2256 |        "      <td>0.0</td>\n",
2257 |        "      <td>0.000000</td>\n",
2258 |        "      <td>0.000562</td>\n",
2259 |        "      <td>17.82</td>\n",
2260 |        "      <td>...</td>\n",
2261 |        "      <td>0</td>\n",
2262 |        "      <td>1</td>\n",
2263 |        "      <td>0</td>\n",
2264 |        "      <td>0</td>\n",
2265 |        "      <td>1</td>\n",
2266 |        "      <td>0</td>\n",
2267 |        "      <td>0</td>\n",
2268 |        "      <td>1</td>\n",
2269 |        "      <td>0</td>\n",
2270 |        "      <td>0</td>\n",
2271 |        "    </tr>\n",
2272 |        "    <tr>\n",
2273 |        "      <th>3</th>\n",
2274 |        "      <td>-0.550912</td>\n",
2275 |        "      <td>29359.21</td>\n",
2276 |        "      <td>0.00</td>\n",
2277 |        "      <td>0.00</td>\n",
2278 |        "      <td>0.000000</td>\n",
2279 |        "      <td>0.0</td>\n",
2280 |        "      <td>0.0</td>\n",
2281 |        "      <td>0.000000</td>\n",
2282 |        "      <td>-0.550912</td>\n",
2283 |        "      <td>33273.57</td>\n",
2284 |        "      <td>...</td>\n",
2285 |        "      <td>0</td>\n",
2286 |        "      <td>1</td>\n",
2287 |        "      <td>0</td>\n",
2288 |        "      <td>0</td>\n",
2289 |        "      <td>1</td>\n",
2290 |        "      <td>0</td>\n",
2291 |        "      <td>0</td>\n",
2292 |        "      <td>0</td>\n",
2293 |        "      <td>1</td>\n",
2294 |        "      <td>0</td>\n",
2295 |        "    </tr>\n",
2296 |        "    <tr>\n",
2297 |        "      <th>4</th>\n",
2298 |        "      <td>0.000288</td>\n",
2299 |        "      <td>34.68</td>\n",
2300 |        "      <td>0.00</td>\n",
2301 |        "      <td>0.00</td>\n",
2302 |        "      <td>0.000000</td>\n",
2303 |        "      <td>0.0</td>\n",
2304 |        "      <td>0.0</td>\n",
2305 |        "      <td>0.000000</td>\n",
2306 |        "      <td>0.000288</td>\n",
2307 |        "      <td>34.70</td>\n",
2308 |        "      <td>...</td>\n",
2309 |        "      <td>0</td>\n",
2310 |        "      <td>1</td>\n",
2311 |        "      <td>0</td>\n",
2312 |        "      <td>0</td>\n",
2313 |        "      <td>0</td>\n",
2314 |        "      <td>1</td>\n",
2315 |        "      <td>0</td>\n",
2316 |        "      <td>1</td>\n",
2317 |        "      <td>0</td>\n",
2318 |        "      <td>0</td>\n",
2319 |        "    </tr>\n",
2320 |        "  </tbody>\n",
2321 |        "</table>\n",
2322 |        "<p>5 rows × 178 columns</p>\n",
2323 |        "</div>"
2324 |       ],
2325 |       "text/plain": [
2326 |        "   LOCAL_CUR_SAV_SLOPE  LOCAL_CUR_MON_AVG_BAL  LOCAL_OVEONEYR_FF_MON_AVG_BAL  \\\n",
2327 |        "0             0.000000                   0.00                       50429.68   \n",
2328 |        "1             1.005692               20149.04                       56047.50   \n",
2329 |        "2             0.000562                  17.81                           0.00   \n",
2330 |        "3            -0.550912               29359.21                           0.00   \n",
2331 |        "4             0.000288                  34.68                           0.00   \n",
2332 |        "\n",
2333 |        "   LOCAL_FIX_MON_AVG_BAL  LOCAL_FIX_MON_AVG_BAL_PROP  LOCAL_BELONEYR_FF_SLOPE  \\\n",
2334 |        "0               50429.68                    1.000000                      0.0   \n",
2335 |        "1               56047.50                    0.735565                      0.0   \n",
2336 |        "2                   0.00                    0.000000                      0.0   \n",
2337 |        "3                   0.00                    0.000000                      0.0   \n",
2338 |        "4                   0.00                    0.000000                      0.0   \n",
2339 |        "\n",
2340 |        "   LOCAL_BELONEYR_FF_MON_AVG_BAL  LOCAL_OVEONEYR_FF_SLOPE  LOCAL_SAV_SLOPE  \\\n",
2341 |        "0                            0.0                -0.032395        -0.032395   \n",
2342 |        "1                            0.0                -0.148857         0.003965   \n",
2343 |        "2                            0.0                 0.000000         0.000562   \n",
2344 |        "3                            0.0                 0.000000        -0.550912   \n",
2345 |        "4                            0.0                 0.000000         0.000288   \n",
2346 |        "\n",
2347 |        "   LOCAL_SAV_CUR_ALL_BAL     ...       kid6_10_nan  kid11_15_U  kid11_15_Y  \\\n",
2348 |        "0               41000.00     ...                 0           1           0   \n",
2349 |        "1               31929.10     ...                 0           1           0   \n",
2350 |        "2                  17.82     ...                 0           1           0   \n",
2351 |        "3               33273.57     ...                 0           1           0   \n",
2352 |        "4                  34.70     ...                 0           1           0   \n",
2353 |        "\n",
2354 |        "   kid11_15_nan  kid16_17_U  kid16_17_Y  kid16_17_nan  car_buy_New  \\\n",
2355 |        "0             0           1           0             0            1   \n",
2356 |        "1             0           1           0             0            1   \n",
2357 |        "2             0           1           0             0            1   \n",
2358 |        "3             0           1           0             0            0   \n",
2359 |        "4             0           0           1             0            1   \n",
2360 |        "\n",
2361 |        "   car_buy_UNKNOWN  car_buy_nan  \n",
2362 |        "0                0            0  \n",
2363 |        "1                0            0  \n",
2364 |        "2                0            0  \n",
2365 |        "3                1            0  \n",
2366 |        "4                0            0  \n",
2367 |        "\n",
2368 |        "[5 rows x 178 columns]"
2369 |       ]
2370 |      },
2371 |      "execution_count": 20,
2372 |      "metadata": {},
2373 |      "output_type": "execute_result"
2374 |     }
2375 |    ],
2376 |    "source": [
2377 |     "model_data.head()"
2378 |    ]
2379 |   },
2380 |   {
2381 |    "cell_type": "code",
2382 |    "execution_count": 21,
2383 |    "metadata": {},
2384 |    "outputs": [
2385 |     {
2386 |      "data": {
2387 |       "text/plain": [
2388 |        "(17241, 178)"
2389 |       ]
2390 |      },
2391 |      "execution_count": 21,
2392 |      "metadata": {},
2393 |      "output_type": "execute_result"
2394 |     }
2395 |    ],
2396 |    "source": [
2397 |     "model_data.shape"
2398 |    ]
2399 |   },
2400 |   {
2401 |    "cell_type": "code",
2402 |    "execution_count": 22,
2403 |    "metadata": {
2404 |     "collapsed": true
2405 |    },
2406 |    "outputs": [],
2407 |    "source": [
2408 |     "# 处理各个特征取值的范围不一致性\n",
2409 |     "# 将全部特征的取值均限制在[0,1]之间\n",
2410 |     "model_data = (model_data - model_data.min()) / (model_data.max() - model_data.min())\n",
2411 |     "model_data.replace(float('inf'), 1, inplace=True)\n",
2412 |     "model_data.fillna(0, inplace=True)"
2413 |    ]
2414 |   },
2415 |   {
2416 |    "cell_type": "code",
2417 |    "execution_count": 23,
2418 |    "metadata": {},
2419 |    "outputs": [
2420 |     {
2421 |      "name": "stdout",
2422 |      "output_type": "stream",
2423 |      "text": [
2424 |       "False\n",
2425 |       "False\n",
2426 |       "False\n",
2427 |       "False\n",
2428 |       "False\n",
2429 |       "False\n",
2430 |       "False\n",
2431 |       "False\n",
2432 |       "False\n",
2433 |       "False\n",
2434 |       "False\n",
2435 |       "False\n",
2436 |       "False\n",
2437 |       "False\n",
2438 |       "False\n",
2439 |       "False\n",
2440 |       "False\n",
2441 |       "False\n",
2442 |       "False\n",
2443 |       "False\n",
2444 |       "False\n",
2445 |       "False\n",
2446 |       "False\n",
2447 |       "False\n",
2448 |       "False\n",
2449 |       "False\n",
2450 |       "False\n",
2451 |       "False\n",
2452 |       "False\n",
2453 |       "False\n",
2454 |       "False\n",
2455 |       "False\n",
2456 |       "False\n",
2457 |       "False\n",
2458 |       "False\n",
2459 |       "False\n",
2460 |       "False\n",
2461 |       "False\n",
2462 |       "False\n",
2463 |       "False\n",
2464 |       "False\n",
2465 |       "False\n",
2466 |       "False\n",
2467 |       "False\n",
2468 |       "False\n",
2469 |       "False\n",
2470 |       "False\n",
2471 |       "False\n",
2472 |       "False\n",
2473 |       "False\n",
2474 |       "False\n",
2475 |       "False\n",
2476 |       "False\n",
2477 |       "False\n",
2478 |       "False\n",
2479 |       "False\n",
2480 |       "False\n",
2481 |       "False\n",
2482 |       "False\n",
2483 |       "False\n",
2484 |       "False\n",
2485 |       "False\n",
2486 |       "False\n",
2487 |       "False\n",
2488 |       "False\n",
2489 |       "False\n",
2490 |       "False\n",
2491 |       "False\n",
2492 |       "False\n",
2493 |       "False\n",
2494 |       "False\n",
2495 |       "False\n",
2496 |       "False\n",
2497 |       "False\n",
2498 |       "False\n",
2499 |       "False\n",
2500 |       "False\n",
2501 |       "False\n",
2502 |       "False\n",
2503 |       "False\n",
2504 |       "False\n",
2505 |       "False\n",
2506 |       "False\n",
2507 |       "False\n",
2508 |       "False\n",
2509 |       "False\n",
2510 |       "False\n",
2511 |       "False\n",
2512 |       "False\n",
2513 |       "False\n",
2514 |       "False\n",
2515 |       "False\n",
2516 |       "False\n",
2517 |       "False\n",
2518 |       "False\n",
2519 |       "False\n",
2520 |       "False\n",
2521 |       "False\n",
2522 |       "False\n",
2523 |       "False\n",
2524 |       "False\n",
2525 |       "False\n",
2526 |       "False\n",
2527 |       "False\n",
2528 |       "False\n",
2529 |       "False\n",
2530 |       "False\n",
2531 |       "False\n",
2532 |       "False\n",
2533 |       "False\n",
2534 |       "False\n",
2535 |       "False\n",
2536 |       "False\n",
2537 |       "False\n",
2538 |       "False\n",
2539 |       "False\n",
2540 |       "False\n",
2541 |       "False\n",
2542 |       "False\n",
2543 |       "False\n",
2544 |       "False\n",
2545 |       "False\n",
2546 |       "False\n",
2547 |       "False\n",
2548 |       "False\n",
2549 |       "False\n",
2550 |       "False\n",
2551 |       "False\n",
2552 |       "False\n",
2553 |       "False\n",
2554 |       "False\n",
2555 |       "False\n",
2556 |       "False\n",
2557 |       "False\n",
2558 |       "False\n",
2559 |       "False\n",
2560 |       "False\n",
2561 |       "False\n",
2562 |       "False\n",
2563 |       "False\n",
2564 |       "False\n",
2565 |       "False\n",
2566 |       "False\n",
2567 |       "False\n",
2568 |       "False\n",
2569 |       "False\n",
2570 |       "False\n",
2571 |       "False\n",
2572 |       "False\n",
2573 |       "False\n",
2574 |       "False\n",
2575 |       "False\n",
2576 |       "False\n",
2577 |       "False\n",
2578 |       "False\n",
2579 |       "False\n",
2580 |       "False\n",
2581 |       "False\n",
2582 |       "False\n",
2583 |       "False\n",
2584 |       "False\n",
2585 |       "False\n",
2586 |       "False\n",
2587 |       "False\n",
2588 |       "False\n",
2589 |       "False\n",
2590 |       "False\n",
2591 |       "False\n",
2592 |       "False\n",
2593 |       "False\n",
2594 |       "False\n",
2595 |       "False\n",
2596 |       "False\n",
2597 |       "False\n",
2598 |       "False\n",
2599 |       "False\n",
2600 |       "False\n",
2601 |       "False\n"
2602 |      ]
2603 |     }
2604 |    ],
2605 |    "source": [
2606 |     "null_test = model_data.isnull().any()\n",
2607 |     "for i in null_test:\n",
2608 |     "    print(i)"
2609 |    ]
2610 |   },
2611 |   {
2612 |    "cell_type": "code",
2613 |    "execution_count": 24,
2614 |    "metadata": {
2615 |     "collapsed": true
2616 |    },
2617 |    "outputs": [],
2618 |    "source": [
2619 |     "model_data.to_csv('./model_data.csv', encoding='utf-8', index=False)"
2620 |    ]
2621 |   },
2622 |   {
2623 |    "cell_type": "code",
2624 |    "execution_count": 25,
2625 |    "metadata": {},
2626 |    "outputs": [
2627 |     {
2628 |      "data": {
2629 |       "text/plain": [
2630 |        "0.0    15500\n",
2631 |        "1.0     1741\n",
2632 |        "Name: CHURN_CUST_IND, dtype: int64"
2633 |       ]
2634 |      },
2635 |      "execution_count": 25,
2636 |      "metadata": {},
2637 |      "output_type": "execute_result"
2638 |     }
2639 |    ],
2640 |    "source": [
2641 |     "model_data['CHURN_CUST_IND'].value_counts()"
2642 |    ]
2643 |   },
2644 |   {
2645 |    "cell_type": "code",
2646 |    "execution_count": 26,
2647 |    "metadata": {
2648 |     "collapsed": true
2649 |    },
2650 |    "outputs": [],
2651 |    "source": [
2652 |     "model_data_des = model_data.describe().T"
2653 |    ]
2654 |   },
2655 |   {
2656 |    "cell_type": "code",
2657 |    "execution_count": 27,
2658 |    "metadata": {},
2659 |    "outputs": [
2660 |     {
2661 |      "data": {
2662 |       "text/html": [
2663 |        "<div>\n",
2664 |        "<style>\n",
2665 |        "    .dataframe thead tr:only-child th {\n",
2666 |        "        text-align: right;\n",
2667 |        "    }\n",
2668 |        "\n",
2669 |        "    .dataframe thead th {\n",
2670 |        "        text-align: left;\n",
2671 |        "    }\n",
2672 |        "\n",
2673 |        "    .dataframe tbody tr th {\n",
2674 |        "        vertical-align: top;\n",
2675 |        "    }\n",
2676 |        "</style>\n",
2677 |        "<table border=\"1\" class=\"dataframe\">\n",
2678 |        "  <thead>\n",
2679 |        "    <tr style=\"text-align: right;\">\n",
2680 |        "      <th></th>\n",
2681 |        "      <th>count</th>\n",
2682 |        "      <th>mean</th>\n",
2683 |        "      <th>std</th>\n",
2684 |        "      <th>min</th>\n",
2685 |        "      <th>25%</th>\n",
2686 |        "      <th>50%</th>\n",
2687 |        "      <th>75%</th>\n",
2688 |        "      <th>max</th>\n",
2689 |        "    </tr>\n",
2690 |        "  </thead>\n",
2691 |        "  <tbody>\n",
2692 |        "    <tr>\n",
2693 |        "      <th>LOCAL_CUR_SAV_SLOPE</th>\n",
2694 |        "      <td>17241.0</td>\n",
2695 |        "      <td>0.329458</td>\n",
2696 |        "      <td>0.135781</td>\n",
2697 |        "      <td>0.0</td>\n",
2698 |        "      <td>0.304001</td>\n",
2699 |        "      <td>0.333333</td>\n",
2700 |        "      <td>0.358169</td>\n",
2701 |        "      <td>1.0</td>\n",
2702 |        "    </tr>\n",
2703 |        "    <tr>\n",
2704 |        "      <th>LOCAL_CUR_MON_AVG_BAL</th>\n",
2705 |        "      <td>17241.0</td>\n",
2706 |        "      <td>0.003645</td>\n",
2707 |        "      <td>0.012958</td>\n",
2708 |        "      <td>0.0</td>\n",
2709 |        "      <td>0.000007</td>\n",
2710 |        "      <td>0.000425</td>\n",
2711 |        "      <td>0.003037</td>\n",
2712 |        "      <td>1.0</td>\n",
2713 |        "    </tr>\n",
2714 |        "    <tr>\n",
2715 |        "      <th>LOCAL_OVEONEYR_FF_MON_AVG_BAL</th>\n",
2716 |        "      <td>17241.0</td>\n",
2717 |        "      <td>0.012951</td>\n",
2718 |        "      <td>0.023193</td>\n",
2719 |        "      <td>0.0</td>\n",
2720 |        "      <td>0.000000</td>\n",
2721 |        "      <td>0.008991</td>\n",
2722 |        "      <td>0.017082</td>\n",
2723 |        "      <td>1.0</td>\n",
2724 |        "    </tr>\n",
2725 |        "    <tr>\n",
2726 |        "      <th>LOCAL_FIX_MON_AVG_BAL</th>\n",
2727 |        "      <td>17241.0</td>\n",
2728 |        "      <td>0.001925</td>\n",
2729 |        "      <td>0.008266</td>\n",
2730 |        "      <td>0.0</td>\n",
2731 |        "      <td>0.000000</td>\n",
2732 |        "      <td>0.001300</td>\n",
2733 |        "      <td>0.002320</td>\n",
2734 |        "      <td>1.0</td>\n",
2735 |        "    </tr>\n",
2736 |        "    <tr>\n",
2737 |        "      <th>LOCAL_FIX_MON_AVG_BAL_PROP</th>\n",
2738 |        "      <td>17241.0</td>\n",
2739 |        "      <td>0.587804</td>\n",
2740 |        "      <td>0.443246</td>\n",
2741 |        "      <td>0.0</td>\n",
2742 |        "      <td>0.000000</td>\n",
2743 |        "      <td>0.854396</td>\n",
2744 |        "      <td>0.990436</td>\n",
2745 |        "      <td>1.0</td>\n",
2746 |        "    </tr>\n",
2747 |        "    <tr>\n",
2748 |        "      <th>LOCAL_BELONEYR_FF_SLOPE</th>\n",
2749 |        "      <td>17241.0</td>\n",
2750 |        "      <td>0.350814</td>\n",
2751 |        "      <td>0.088405</td>\n",
2752 |        "      <td>0.0</td>\n",
2753 |        "      <td>0.344437</td>\n",
2754 |        "      <td>0.344437</td>\n",
2755 |        "      <td>0.344437</td>\n",
2756 |        "      <td>1.0</td>\n",
2757 |        "    </tr>\n",
2758 |        "    <tr>\n",
2759 |        "      <th>LOCAL_BELONEYR_FF_MON_AVG_BAL</th>\n",
2760 |        "      <td>17241.0</td>\n",
2761 |        "      <td>0.000271</td>\n",
2762 |        "      <td>0.007734</td>\n",
2763 |        "      <td>0.0</td>\n",
2764 |        "      <td>0.000000</td>\n",
2765 |        "      <td>0.000000</td>\n",
2766 |        "      <td>0.000000</td>\n",
2767 |        "      <td>1.0</td>\n",
2768 |        "    </tr>\n",
2769 |        "    <tr>\n",
2770 |        "      <th>LOCAL_OVEONEYR_FF_SLOPE</th>\n",
2771 |        "      <td>17241.0</td>\n",
2772 |        "      <td>0.350590</td>\n",
2773 |        "      <td>0.070418</td>\n",
2774 |        "      <td>0.0</td>\n",
2775 |        "      <td>0.344442</td>\n",
2776 |        "      <td>0.344442</td>\n",
2777 |        "      <td>0.345026</td>\n",
2778 |        "      <td>1.0</td>\n",
2779 |        "    </tr>\n",
2780 |        "    <tr>\n",
2781 |        "      <th>LOCAL_SAV_SLOPE</th>\n",
2782 |        "      <td>17241.0</td>\n",
2783 |        "      <td>0.342997</td>\n",
2784 |        "      <td>0.088393</td>\n",
2785 |        "      <td>0.0</td>\n",
2786 |        "      <td>0.344164</td>\n",
2787 |        "      <td>0.344535</td>\n",
2788 |        "      <td>0.356539</td>\n",
2789 |        "      <td>1.0</td>\n",
2790 |        "    </tr>\n",
2791 |        "    <tr>\n",
2792 |        "      <th>LOCAL_SAV_CUR_ALL_BAL</th>\n",
2793 |        "      <td>17241.0</td>\n",
2794 |        "      <td>0.002509</td>\n",
2795 |        "      <td>0.008524</td>\n",
2796 |        "      <td>0.0</td>\n",
2797 |        "      <td>0.001131</td>\n",
2798 |        "      <td>0.001710</td>\n",
2799 |        "      <td>0.002807</td>\n",
2800 |        "      <td>1.0</td>\n",
2801 |        "    </tr>\n",
2802 |        "    <tr>\n",
2803 |        "      <th>LOCAL_SAV_MON_AVG_BAL</th>\n",
2804 |        "      <td>17241.0</td>\n",
2805 |        "      <td>0.002562</td>\n",
2806 |        "      <td>0.008540</td>\n",
2807 |        "      <td>0.0</td>\n",
2808 |        "      <td>0.001160</td>\n",
2809 |        "      <td>0.001749</td>\n",
2810 |        "      <td>0.002861</td>\n",
2811 |        "      <td>1.0</td>\n",
2812 |        "    </tr>\n",
2813 |        "    <tr>\n",
2814 |        "      <th>SAV_SLOPE</th>\n",
2815 |        "      <td>17241.0</td>\n",
2816 |        "      <td>0.342959</td>\n",
2817 |        "      <td>0.088283</td>\n",
2818 |        "      <td>0.0</td>\n",
2819 |        "      <td>0.344162</td>\n",
2820 |        "      <td>0.344535</td>\n",
2821 |        "      <td>0.356519</td>\n",
2822 |        "      <td>1.0</td>\n",
2823 |        "    </tr>\n",
2824 |        "    <tr>\n",
2825 |        "      <th>SAV_CUR_ALL_BAL</th>\n",
2826 |        "      <td>17241.0</td>\n",
2827 |        "      <td>0.002511</td>\n",
2828 |        "      <td>0.008525</td>\n",
2829 |        "      <td>0.0</td>\n",
2830 |        "      <td>0.001131</td>\n",
2831 |        "      <td>0.001713</td>\n",
2832 |        "      <td>0.002809</td>\n",
2833 |        "      <td>1.0</td>\n",
2834 |        "    </tr>\n",
2835 |        "    <tr>\n",
2836 |        "      <th>SAV_MON_AVG_BAL</th>\n",
2837 |        "      <td>17241.0</td>\n",
2838 |        "      <td>0.002563</td>\n",
2839 |        "      <td>0.008541</td>\n",
2840 |        "      <td>0.0</td>\n",
2841 |        "      <td>0.001161</td>\n",
2842 |        "      <td>0.001751</td>\n",
2843 |        "      <td>0.002862</td>\n",
2844 |        "      <td>1.0</td>\n",
2845 |        "    </tr>\n",
2846 |        "    <tr>\n",
2847 |        "      <th>FR_SAV_CUR_ALL_BAL</th>\n",
2848 |        "      <td>17241.0</td>\n",
2849 |        "      <td>0.000441</td>\n",
2850 |        "      <td>0.014429</td>\n",
2851 |        "      <td>0.0</td>\n",
2852 |        "      <td>0.000000</td>\n",
2853 |        "      <td>0.000000</td>\n",
2854 |        "      <td>0.000000</td>\n",
2855 |        "      <td>1.0</td>\n",
2856 |        "    </tr>\n",
2857 |        "    <tr>\n",
2858 |        "      <th>ASSET_CUR_ALL_BAL</th>\n",
2859 |        "      <td>17241.0</td>\n",
2860 |        "      <td>0.002650</td>\n",
2861 |        "      <td>0.008598</td>\n",
2862 |        "      <td>0.0</td>\n",
2863 |        "      <td>0.001142</td>\n",
2864 |        "      <td>0.001781</td>\n",
2865 |        "      <td>0.002941</td>\n",
2866 |        "      <td>1.0</td>\n",
2867 |        "    </tr>\n",
2868 |        "    <tr>\n",
2869 |        "      <th>ASSET_MON_AVG_BAL</th>\n",
2870 |        "      <td>17241.0</td>\n",
2871 |        "      <td>0.002701</td>\n",
2872 |        "      <td>0.008613</td>\n",
2873 |        "      <td>0.0</td>\n",
2874 |        "      <td>0.001193</td>\n",
2875 |        "      <td>0.001816</td>\n",
2876 |        "      <td>0.002988</td>\n",
2877 |        "      <td>1.0</td>\n",
2878 |        "    </tr>\n",
2879 |        "    <tr>\n",
2880 |        "      <th>LOCAL_CUR_TRANS_TX_AMT</th>\n",
2881 |        "      <td>17241.0</td>\n",
2882 |        "      <td>0.000417</td>\n",
2883 |        "      <td>0.010761</td>\n",
2884 |        "      <td>0.0</td>\n",
2885 |        "      <td>0.000000</td>\n",
2886 |        "      <td>0.000000</td>\n",
2887 |        "      <td>0.000000</td>\n",
2888 |        "      <td>1.0</td>\n",
2889 |        "    </tr>\n",
2890 |        "    <tr>\n",
2891 |        "      <th>LOCAL_CUR_LASTSAV_TX_AMT</th>\n",
2892 |        "      <td>17241.0</td>\n",
2893 |        "      <td>0.001628</td>\n",
2894 |        "      <td>0.013896</td>\n",
2895 |        "      <td>0.0</td>\n",
2896 |        "      <td>0.000000</td>\n",
2897 |        "      <td>0.000000</td>\n",
2898 |        "      <td>0.000000</td>\n",
2899 |        "      <td>1.0</td>\n",
2900 |        "    </tr>\n",
2901 |        "    <tr>\n",
2902 |        "      <th>LOCAL_CUR_WITHDRAW_TX_AMT</th>\n",
2903 |        "      <td>17241.0</td>\n",
2904 |        "      <td>0.002978</td>\n",
2905 |        "      <td>0.018774</td>\n",
2906 |        "      <td>0.0</td>\n",
2907 |        "      <td>0.000000</td>\n",
2908 |        "      <td>0.000000</td>\n",
2909 |        "      <td>0.000492</td>\n",
2910 |        "      <td>1.0</td>\n",
2911 |        "    </tr>\n",
2912 |        "    <tr>\n",
2913 |        "      <th>LOCAL_FIX_OPEN_ACC_TX_AMT</th>\n",
2914 |        "      <td>17241.0</td>\n",
2915 |        "      <td>0.002138</td>\n",
2916 |        "      <td>0.013118</td>\n",
2917 |        "      <td>0.0</td>\n",
2918 |        "      <td>0.000000</td>\n",
2919 |        "      <td>0.000000</td>\n",
2920 |        "      <td>0.000000</td>\n",
2921 |        "      <td>1.0</td>\n",
2922 |        "    </tr>\n",
2923 |        "    <tr>\n",
2924 |        "      <th>LOCAL_FIX_WITHDRAW_TX_AMT</th>\n",
2925 |        "      <td>17241.0</td>\n",
2926 |        "      <td>0.001221</td>\n",
2927 |        "      <td>0.013362</td>\n",
2928 |        "      <td>0.0</td>\n",
2929 |        "      <td>0.000000</td>\n",
2930 |        "      <td>0.000000</td>\n",
2931 |        "      <td>0.000000</td>\n",
2932 |        "      <td>1.0</td>\n",
2933 |        "    </tr>\n",
2934 |        "    <tr>\n",
2935 |        "      <th>LOCAL_FIX_CLOSE_ACC_TX_AMT</th>\n",
2936 |        "      <td>17241.0</td>\n",
2937 |        "      <td>0.004347</td>\n",
2938 |        "      <td>0.026010</td>\n",
2939 |        "      <td>0.0</td>\n",
2940 |        "      <td>0.000000</td>\n",
2941 |        "      <td>0.000000</td>\n",
2942 |        "      <td>0.000000</td>\n",
2943 |        "      <td>1.0</td>\n",
2944 |        "    </tr>\n",
2945 |        "    <tr>\n",
2946 |        "      <th>L6M_INDFINA_ALL_TX_AMT</th>\n",
2947 |        "      <td>17241.0</td>\n",
2948 |        "      <td>0.001079</td>\n",
2949 |        "      <td>0.019296</td>\n",
2950 |        "      <td>0.0</td>\n",
2951 |        "      <td>0.000000</td>\n",
2952 |        "      <td>0.000000</td>\n",
2953 |        "      <td>0.000000</td>\n",
2954 |        "      <td>1.0</td>\n",
2955 |        "    </tr>\n",
2956 |        "    <tr>\n",
2957 |        "      <th>da_Mean</th>\n",
2958 |        "      <td>17241.0</td>\n",
2959 |        "      <td>0.015937</td>\n",
2960 |        "      <td>0.036701</td>\n",
2961 |        "      <td>0.0</td>\n",
2962 |        "      <td>0.000000</td>\n",
2963 |        "      <td>0.004292</td>\n",
2964 |        "      <td>0.017167</td>\n",
2965 |        "      <td>1.0</td>\n",
2966 |        "    </tr>\n",
2967 |        "    <tr>\n",
2968 |        "      <th>ovrmou_Mean</th>\n",
2969 |        "      <td>17241.0</td>\n",
2970 |        "      <td>0.021053</td>\n",
2971 |        "      <td>0.049172</td>\n",
2972 |        "      <td>0.0</td>\n",
2973 |        "      <td>0.000000</td>\n",
2974 |        "      <td>0.001457</td>\n",
2975 |        "      <td>0.021592</td>\n",
2976 |        "      <td>1.0</td>\n",
2977 |        "    </tr>\n",
2978 |        "    <tr>\n",
2979 |        "      <th>ovrrev_Mean</th>\n",
2980 |        "      <td>17241.0</td>\n",
2981 |        "      <td>0.021747</td>\n",
2982 |        "      <td>0.049016</td>\n",
2983 |        "      <td>0.0</td>\n",
2984 |        "      <td>0.000000</td>\n",
2985 |        "      <td>0.001622</td>\n",
2986 |        "      <td>0.022766</td>\n",
2987 |        "      <td>1.0</td>\n",
2988 |        "    </tr>\n",
2989 |        "    <tr>\n",
2990 |        "      <th>vceovr_Mean</th>\n",
2991 |        "      <td>17241.0</td>\n",
2992 |        "      <td>0.021327</td>\n",
2993 |        "      <td>0.048654</td>\n",
2994 |        "      <td>0.0</td>\n",
2995 |        "      <td>0.000000</td>\n",
2996 |        "      <td>0.000873</td>\n",
2997 |        "      <td>0.021976</td>\n",
2998 |        "      <td>1.0</td>\n",
2999 |        "    </tr>\n",
3000 |        "    <tr>\n",
3001 |        "      <th>datovr_Mean</th>\n",
3002 |        "      <td>17241.0</td>\n",
3003 |        "      <td>0.002470</td>\n",
3004 |        "      <td>0.021156</td>\n",
3005 |        "      <td>0.0</td>\n",
3006 |        "      <td>0.000000</td>\n",
3007 |        "      <td>0.000000</td>\n",
3008 |        "      <td>0.000000</td>\n",
3009 |        "      <td>1.0</td>\n",
3010 |        "    </tr>\n",
3011 |        "    <tr>\n",
3012 |        "      <th>roam_Mean</th>\n",
3013 |        "      <td>17241.0</td>\n",
3014 |        "      <td>0.002738</td>\n",
3015 |        "      <td>0.015707</td>\n",
3016 |        "      <td>0.0</td>\n",
3017 |        "      <td>0.000000</td>\n",
3018 |        "      <td>0.000000</td>\n",
3019 |        "      <td>0.000619</td>\n",
3020 |        "      <td>1.0</td>\n",
3021 |        "    </tr>\n",
3022 |        "    <tr>\n",
3023 |        "      <th>...</th>\n",
3024 |        "      <td>...</td>\n",
3025 |        "      <td>...</td>\n",
3026 |        "      <td>...</td>\n",
3027 |        "      <td>...</td>\n",
3028 |        "      <td>...</td>\n",
3029 |        "      <td>...</td>\n",
3030 |        "      <td>...</td>\n",
3031 |        "      <td>...</td>\n",
3032 |        "    </tr>\n",
3033 |        "    <tr>\n",
3034 |        "      <th>marital_nan</th>\n",
3035 |        "      <td>17241.0</td>\n",
3036 |        "      <td>0.017748</td>\n",
3037 |        "      <td>0.132039</td>\n",
3038 |        "      <td>0.0</td>\n",
3039 |        "      <td>0.000000</td>\n",
3040 |        "      <td>0.000000</td>\n",
3041 |        "      <td>0.000000</td>\n",
3042 |        "      <td>1.0</td>\n",
3043 |        "    </tr>\n",
3044 |        "    <tr>\n",
3045 |        "      <th>wrkwoman_Y</th>\n",
3046 |        "      <td>17241.0</td>\n",
3047 |        "      <td>0.129343</td>\n",
3048 |        "      <td>0.335589</td>\n",
3049 |        "      <td>0.0</td>\n",
3050 |        "      <td>0.000000</td>\n",
3051 |        "      <td>0.000000</td>\n",
3052 |        "      <td>0.000000</td>\n",
3053 |        "      <td>1.0</td>\n",
3054 |        "    </tr>\n",
3055 |        "    <tr>\n",
3056 |        "      <th>wrkwoman_nan</th>\n",
3057 |        "      <td>17241.0</td>\n",
3058 |        "      <td>0.870657</td>\n",
3059 |        "      <td>0.335589</td>\n",
3060 |        "      <td>0.0</td>\n",
3061 |        "      <td>1.000000</td>\n",
3062 |        "      <td>1.000000</td>\n",
3063 |        "      <td>1.000000</td>\n",
3064 |        "      <td>1.0</td>\n",
3065 |        "    </tr>\n",
3066 |        "    <tr>\n",
3067 |        "      <th>proptype_A</th>\n",
3068 |        "      <td>17241.0</td>\n",
3069 |        "      <td>0.259092</td>\n",
3070 |        "      <td>0.438149</td>\n",
3071 |        "      <td>0.0</td>\n",
3072 |        "      <td>0.000000</td>\n",
3073 |        "      <td>0.000000</td>\n",
3074 |        "      <td>1.000000</td>\n",
3075 |        "      <td>1.0</td>\n",
3076 |        "    </tr>\n",
3077 |        "    <tr>\n",
3078 |        "      <th>proptype_B</th>\n",
3079 |        "      <td>17241.0</td>\n",
3080 |        "      <td>0.015370</td>\n",
3081 |        "      <td>0.123024</td>\n",
3082 |        "      <td>0.0</td>\n",
3083 |        "      <td>0.000000</td>\n",
3084 |        "      <td>0.000000</td>\n",
3085 |        "      <td>0.000000</td>\n",
3086 |        "      <td>1.0</td>\n",
3087 |        "    </tr>\n",
3088 |        "    <tr>\n",
3089 |        "      <th>proptype_D</th>\n",
3090 |        "      <td>17241.0</td>\n",
3091 |        "      <td>0.006902</td>\n",
3092 |        "      <td>0.082794</td>\n",
3093 |        "      <td>0.0</td>\n",
3094 |        "      <td>0.000000</td>\n",
3095 |        "      <td>0.000000</td>\n",
3096 |        "      <td>0.000000</td>\n",
3097 |        "      <td>1.0</td>\n",
3098 |        "    </tr>\n",
3099 |        "    <tr>\n",
3100 |        "      <th>proptype_E</th>\n",
3101 |        "      <td>17241.0</td>\n",
3102 |        "      <td>0.003770</td>\n",
3103 |        "      <td>0.061287</td>\n",
3104 |        "      <td>0.0</td>\n",
3105 |        "      <td>0.000000</td>\n",
3106 |        "      <td>0.000000</td>\n",
3107 |        "      <td>0.000000</td>\n",
3108 |        "      <td>1.0</td>\n",
3109 |        "    </tr>\n",
3110 |        "    <tr>\n",
3111 |        "      <th>proptype_G</th>\n",
3112 |        "      <td>17241.0</td>\n",
3113 |        "      <td>0.000870</td>\n",
3114 |        "      <td>0.029484</td>\n",
3115 |        "      <td>0.0</td>\n",
3116 |        "      <td>0.000000</td>\n",
3117 |        "      <td>0.000000</td>\n",
3118 |        "      <td>0.000000</td>\n",
3119 |        "      <td>1.0</td>\n",
3120 |        "    </tr>\n",
3121 |        "    <tr>\n",
3122 |        "      <th>proptype_M</th>\n",
3123 |        "      <td>17241.0</td>\n",
3124 |        "      <td>0.001972</td>\n",
3125 |        "      <td>0.044365</td>\n",
3126 |        "      <td>0.0</td>\n",
3127 |        "      <td>0.000000</td>\n",
3128 |        "      <td>0.000000</td>\n",
3129 |        "      <td>0.000000</td>\n",
3130 |        "      <td>1.0</td>\n",
3131 |        "    </tr>\n",
3132 |        "    <tr>\n",
3133 |        "      <th>proptype_nan</th>\n",
3134 |        "      <td>17241.0</td>\n",
3135 |        "      <td>0.712024</td>\n",
3136 |        "      <td>0.452833</td>\n",
3137 |        "      <td>0.0</td>\n",
3138 |        "      <td>0.000000</td>\n",
3139 |        "      <td>1.000000</td>\n",
3140 |        "      <td>1.000000</td>\n",
3141 |        "      <td>1.0</td>\n",
3142 |        "    </tr>\n",
3143 |        "    <tr>\n",
3144 |        "      <th>pcowner_Y</th>\n",
3145 |        "      <td>17241.0</td>\n",
3146 |        "      <td>0.187460</td>\n",
3147 |        "      <td>0.390292</td>\n",
3148 |        "      <td>0.0</td>\n",
3149 |        "      <td>0.000000</td>\n",
3150 |        "      <td>0.000000</td>\n",
3151 |        "      <td>0.000000</td>\n",
3152 |        "      <td>1.0</td>\n",
3153 |        "    </tr>\n",
3154 |        "    <tr>\n",
3155 |        "      <th>pcowner_nan</th>\n",
3156 |        "      <td>17241.0</td>\n",
3157 |        "      <td>0.812540</td>\n",
3158 |        "      <td>0.390292</td>\n",
3159 |        "      <td>0.0</td>\n",
3160 |        "      <td>1.000000</td>\n",
3161 |        "      <td>1.000000</td>\n",
3162 |        "      <td>1.000000</td>\n",
3163 |        "      <td>1.0</td>\n",
3164 |        "    </tr>\n",
3165 |        "    <tr>\n",
3166 |        "      <th>kid0_2_U</th>\n",
3167 |        "      <td>17241.0</td>\n",
3168 |        "      <td>0.943623</td>\n",
3169 |        "      <td>0.230655</td>\n",
3170 |        "      <td>0.0</td>\n",
3171 |        "      <td>1.000000</td>\n",
3172 |        "      <td>1.000000</td>\n",
3173 |        "      <td>1.000000</td>\n",
3174 |        "      <td>1.0</td>\n",
3175 |        "    </tr>\n",
3176 |        "    <tr>\n",
3177 |        "      <th>kid0_2_Y</th>\n",
3178 |        "      <td>17241.0</td>\n",
3179 |        "      <td>0.038629</td>\n",
3180 |        "      <td>0.192714</td>\n",
3181 |        "      <td>0.0</td>\n",
3182 |        "      <td>0.000000</td>\n",
3183 |        "      <td>0.000000</td>\n",
3184 |        "      <td>0.000000</td>\n",
3185 |        "      <td>1.0</td>\n",
3186 |        "    </tr>\n",
3187 |        "    <tr>\n",
3188 |        "      <th>kid0_2_nan</th>\n",
3189 |        "      <td>17241.0</td>\n",
3190 |        "      <td>0.017748</td>\n",
3191 |        "      <td>0.132039</td>\n",
3192 |        "      <td>0.0</td>\n",
3193 |        "      <td>0.000000</td>\n",
3194 |        "      <td>0.000000</td>\n",
3195 |        "      <td>0.000000</td>\n",
3196 |        "      <td>1.0</td>\n",
3197 |        "    </tr>\n",
3198 |        "    <tr>\n",
3199 |        "      <th>kid3_5_U</th>\n",
3200 |        "      <td>17241.0</td>\n",
3201 |        "      <td>0.937591</td>\n",
3202 |        "      <td>0.241905</td>\n",
3203 |        "      <td>0.0</td>\n",
3204 |        "      <td>1.000000</td>\n",
3205 |        "      <td>1.000000</td>\n",
3206 |        "      <td>1.000000</td>\n",
3207 |        "      <td>1.0</td>\n",
3208 |        "    </tr>\n",
3209 |        "    <tr>\n",
3210 |        "      <th>kid3_5_Y</th>\n",
3211 |        "      <td>17241.0</td>\n",
3212 |        "      <td>0.044661</td>\n",
3213 |        "      <td>0.206564</td>\n",
3214 |        "      <td>0.0</td>\n",
3215 |        "      <td>0.000000</td>\n",
3216 |        "      <td>0.000000</td>\n",
3217 |        "      <td>0.000000</td>\n",
3218 |        "      <td>1.0</td>\n",
3219 |        "    </tr>\n",
3220 |        "    <tr>\n",
3221 |        "      <th>kid3_5_nan</th>\n",
3222 |        "      <td>17241.0</td>\n",
3223 |        "      <td>0.017748</td>\n",
3224 |        "      <td>0.132039</td>\n",
3225 |        "      <td>0.0</td>\n",
3226 |        "      <td>0.000000</td>\n",
3227 |        "      <td>0.000000</td>\n",
3228 |        "      <td>0.000000</td>\n",
3229 |        "      <td>1.0</td>\n",
3230 |        "    </tr>\n",
3231 |        "    <tr>\n",
3232 |        "      <th>kid6_10_U</th>\n",
3233 |        "      <td>17241.0</td>\n",
3234 |        "      <td>0.899020</td>\n",
3235 |        "      <td>0.301311</td>\n",
3236 |        "      <td>0.0</td>\n",
3237 |        "      <td>1.000000</td>\n",
3238 |        "      <td>1.000000</td>\n",
3239 |        "      <td>1.000000</td>\n",
3240 |        "      <td>1.0</td>\n",
3241 |        "    </tr>\n",
3242 |        "    <tr>\n",
3243 |        "      <th>kid6_10_Y</th>\n",
3244 |        "      <td>17241.0</td>\n",
3245 |        "      <td>0.083232</td>\n",
3246 |        "      <td>0.276240</td>\n",
3247 |        "      <td>0.0</td>\n",
3248 |        "      <td>0.000000</td>\n",
3249 |        "      <td>0.000000</td>\n",
3250 |        "      <td>0.000000</td>\n",
3251 |        "      <td>1.0</td>\n",
3252 |        "    </tr>\n",
3253 |        "    <tr>\n",
3254 |        "      <th>kid6_10_nan</th>\n",
3255 |        "      <td>17241.0</td>\n",
3256 |        "      <td>0.017748</td>\n",
3257 |        "      <td>0.132039</td>\n",
3258 |        "      <td>0.0</td>\n",
3259 |        "      <td>0.000000</td>\n",
3260 |        "      <td>0.000000</td>\n",
3261 |        "      <td>0.000000</td>\n",
3262 |        "      <td>1.0</td>\n",
3263 |        "    </tr>\n",
3264 |        "    <tr>\n",
3265 |        "      <th>kid11_15_U</th>\n",
3266 |        "      <td>17241.0</td>\n",
3267 |        "      <td>0.891944</td>\n",
3268 |        "      <td>0.310461</td>\n",
3269 |        "      <td>0.0</td>\n",
3270 |        "      <td>1.000000</td>\n",
3271 |        "      <td>1.000000</td>\n",
3272 |        "      <td>1.000000</td>\n",
3273 |        "      <td>1.0</td>\n",
3274 |        "    </tr>\n",
3275 |        "    <tr>\n",
3276 |        "      <th>kid11_15_Y</th>\n",
3277 |        "      <td>17241.0</td>\n",
3278 |        "      <td>0.090308</td>\n",
3279 |        "      <td>0.286631</td>\n",
3280 |        "      <td>0.0</td>\n",
3281 |        "      <td>0.000000</td>\n",
3282 |        "      <td>0.000000</td>\n",
3283 |        "      <td>0.000000</td>\n",
3284 |        "      <td>1.0</td>\n",
3285 |        "    </tr>\n",
3286 |        "    <tr>\n",
3287 |        "      <th>kid11_15_nan</th>\n",
3288 |        "      <td>17241.0</td>\n",
3289 |        "      <td>0.017748</td>\n",
3290 |        "      <td>0.132039</td>\n",
3291 |        "      <td>0.0</td>\n",
3292 |        "      <td>0.000000</td>\n",
3293 |        "      <td>0.000000</td>\n",
3294 |        "      <td>0.000000</td>\n",
3295 |        "      <td>1.0</td>\n",
3296 |        "    </tr>\n",
3297 |        "    <tr>\n",
3298 |        "      <th>kid16_17_U</th>\n",
3299 |        "      <td>17241.0</td>\n",
3300 |        "      <td>0.886085</td>\n",
3301 |        "      <td>0.317717</td>\n",
3302 |        "      <td>0.0</td>\n",
3303 |        "      <td>1.000000</td>\n",
3304 |        "      <td>1.000000</td>\n",
3305 |        "      <td>1.000000</td>\n",
3306 |        "      <td>1.0</td>\n",
3307 |        "    </tr>\n",
3308 |        "    <tr>\n",
3309 |        "      <th>kid16_17_Y</th>\n",
3310 |        "      <td>17241.0</td>\n",
3311 |        "      <td>0.096166</td>\n",
3312 |        "      <td>0.294827</td>\n",
3313 |        "      <td>0.0</td>\n",
3314 |        "      <td>0.000000</td>\n",
3315 |        "      <td>0.000000</td>\n",
3316 |        "      <td>0.000000</td>\n",
3317 |        "      <td>1.0</td>\n",
3318 |        "    </tr>\n",
3319 |        "    <tr>\n",
3320 |        "      <th>kid16_17_nan</th>\n",
3321 |        "      <td>17241.0</td>\n",
3322 |        "      <td>0.017748</td>\n",
3323 |        "      <td>0.132039</td>\n",
3324 |        "      <td>0.0</td>\n",
3325 |        "      <td>0.000000</td>\n",
3326 |        "      <td>0.000000</td>\n",
3327 |        "      <td>0.000000</td>\n",
3328 |        "      <td>1.0</td>\n",
3329 |        "    </tr>\n",
3330 |        "    <tr>\n",
3331 |        "      <th>car_buy_New</th>\n",
3332 |        "      <td>17241.0</td>\n",
3333 |        "      <td>0.422017</td>\n",
3334 |        "      <td>0.493896</td>\n",
3335 |        "      <td>0.0</td>\n",
3336 |        "      <td>0.000000</td>\n",
3337 |        "      <td>0.000000</td>\n",
3338 |        "      <td>1.000000</td>\n",
3339 |        "      <td>1.0</td>\n",
3340 |        "    </tr>\n",
3341 |        "    <tr>\n",
3342 |        "      <th>car_buy_UNKNOWN</th>\n",
3343 |        "      <td>17241.0</td>\n",
3344 |        "      <td>0.560234</td>\n",
3345 |        "      <td>0.496373</td>\n",
3346 |        "      <td>0.0</td>\n",
3347 |        "      <td>0.000000</td>\n",
3348 |        "      <td>1.000000</td>\n",
3349 |        "      <td>1.000000</td>\n",
3350 |        "      <td>1.0</td>\n",
3351 |        "    </tr>\n",
3352 |        "    <tr>\n",
3353 |        "      <th>car_buy_nan</th>\n",
3354 |        "      <td>17241.0</td>\n",
3355 |        "      <td>0.017748</td>\n",
3356 |        "      <td>0.132039</td>\n",
3357 |        "      <td>0.0</td>\n",
3358 |        "      <td>0.000000</td>\n",
3359 |        "      <td>0.000000</td>\n",
3360 |        "      <td>0.000000</td>\n",
3361 |        "      <td>1.0</td>\n",
3362 |        "    </tr>\n",
3363 |        "  </tbody>\n",
3364 |        "</table>\n",
3365 |        "<p>178 rows × 8 columns</p>\n",
3366 |        "</div>"
3367 |       ],
3368 |       "text/plain": [
3369 |        "                                 count      mean       std  min       25%  \\\n",
3370 |        "LOCAL_CUR_SAV_SLOPE            17241.0  0.329458  0.135781  0.0  0.304001   \n",
3371 |        "LOCAL_CUR_MON_AVG_BAL          17241.0  0.003645  0.012958  0.0  0.000007   \n",
3372 |        "LOCAL_OVEONEYR_FF_MON_AVG_BAL  17241.0  0.012951  0.023193  0.0  0.000000   \n",
3373 |        "LOCAL_FIX_MON_AVG_BAL          17241.0  0.001925  0.008266  0.0  0.000000   \n",
3374 |        "LOCAL_FIX_MON_AVG_BAL_PROP     17241.0  0.587804  0.443246  0.0  0.000000   \n",
3375 |        "LOCAL_BELONEYR_FF_SLOPE        17241.0  0.350814  0.088405  0.0  0.344437   \n",
3376 |        "LOCAL_BELONEYR_FF_MON_AVG_BAL  17241.0  0.000271  0.007734  0.0  0.000000   \n",
3377 |        "LOCAL_OVEONEYR_FF_SLOPE        17241.0  0.350590  0.070418  0.0  0.344442   \n",
3378 |        "LOCAL_SAV_SLOPE                17241.0  0.342997  0.088393  0.0  0.344164   \n",
3379 |        "LOCAL_SAV_CUR_ALL_BAL          17241.0  0.002509  0.008524  0.0  0.001131   \n",
3380 |        "LOCAL_SAV_MON_AVG_BAL          17241.0  0.002562  0.008540  0.0  0.001160   \n",
3381 |        "SAV_SLOPE                      17241.0  0.342959  0.088283  0.0  0.344162   \n",
3382 |        "SAV_CUR_ALL_BAL                17241.0  0.002511  0.008525  0.0  0.001131   \n",
3383 |        "SAV_MON_AVG_BAL                17241.0  0.002563  0.008541  0.0  0.001161   \n",
3384 |        "FR_SAV_CUR_ALL_BAL             17241.0  0.000441  0.014429  0.0  0.000000   \n",
3385 |        "ASSET_CUR_ALL_BAL              17241.0  0.002650  0.008598  0.0  0.001142   \n",
3386 |        "ASSET_MON_AVG_BAL              17241.0  0.002701  0.008613  0.0  0.001193   \n",
3387 |        "LOCAL_CUR_TRANS_TX_AMT         17241.0  0.000417  0.010761  0.0  0.000000   \n",
3388 |        "LOCAL_CUR_LASTSAV_TX_AMT       17241.0  0.001628  0.013896  0.0  0.000000   \n",
3389 |        "LOCAL_CUR_WITHDRAW_TX_AMT      17241.0  0.002978  0.018774  0.0  0.000000   \n",
3390 |        "LOCAL_FIX_OPEN_ACC_TX_AMT      17241.0  0.002138  0.013118  0.0  0.000000   \n",
3391 |        "LOCAL_FIX_WITHDRAW_TX_AMT      17241.0  0.001221  0.013362  0.0  0.000000   \n",
3392 |        "LOCAL_FIX_CLOSE_ACC_TX_AMT     17241.0  0.004347  0.026010  0.0  0.000000   \n",
3393 |        "L6M_INDFINA_ALL_TX_AMT         17241.0  0.001079  0.019296  0.0  0.000000   \n",
3394 |        "da_Mean                        17241.0  0.015937  0.036701  0.0  0.000000   \n",
3395 |        "ovrmou_Mean                    17241.0  0.021053  0.049172  0.0  0.000000   \n",
3396 |        "ovrrev_Mean                    17241.0  0.021747  0.049016  0.0  0.000000   \n",
3397 |        "vceovr_Mean                    17241.0  0.021327  0.048654  0.0  0.000000   \n",
3398 |        "datovr_Mean                    17241.0  0.002470  0.021156  0.0  0.000000   \n",
3399 |        "roam_Mean                      17241.0  0.002738  0.015707  0.0  0.000000   \n",
3400 |        "...                                ...       ...       ...  ...       ...   \n",
3401 |        "marital_nan                    17241.0  0.017748  0.132039  0.0  0.000000   \n",
3402 |        "wrkwoman_Y                     17241.0  0.129343  0.335589  0.0  0.000000   \n",
3403 |        "wrkwoman_nan                   17241.0  0.870657  0.335589  0.0  1.000000   \n",
3404 |        "proptype_A                     17241.0  0.259092  0.438149  0.0  0.000000   \n",
3405 |        "proptype_B                     17241.0  0.015370  0.123024  0.0  0.000000   \n",
3406 |        "proptype_D                     17241.0  0.006902  0.082794  0.0  0.000000   \n",
3407 |        "proptype_E                     17241.0  0.003770  0.061287  0.0  0.000000   \n",
3408 |        "proptype_G                     17241.0  0.000870  0.029484  0.0  0.000000   \n",
3409 |        "proptype_M                     17241.0  0.001972  0.044365  0.0  0.000000   \n",
3410 |        "proptype_nan                   17241.0  0.712024  0.452833  0.0  0.000000   \n",
3411 |        "pcowner_Y                      17241.0  0.187460  0.390292  0.0  0.000000   \n",
3412 |        "pcowner_nan                    17241.0  0.812540  0.390292  0.0  1.000000   \n",
3413 |        "kid0_2_U                       17241.0  0.943623  0.230655  0.0  1.000000   \n",
3414 |        "kid0_2_Y                       17241.0  0.038629  0.192714  0.0  0.000000   \n",
3415 |        "kid0_2_nan                     17241.0  0.017748  0.132039  0.0  0.000000   \n",
3416 |        "kid3_5_U                       17241.0  0.937591  0.241905  0.0  1.000000   \n",
3417 |        "kid3_5_Y                       17241.0  0.044661  0.206564  0.0  0.000000   \n",
3418 |        "kid3_5_nan                     17241.0  0.017748  0.132039  0.0  0.000000   \n",
3419 |        "kid6_10_U                      17241.0  0.899020  0.301311  0.0  1.000000   \n",
3420 |        "kid6_10_Y                      17241.0  0.083232  0.276240  0.0  0.000000   \n",
3421 |        "kid6_10_nan                    17241.0  0.017748  0.132039  0.0  0.000000   \n",
3422 |        "kid11_15_U                     17241.0  0.891944  0.310461  0.0  1.000000   \n",
3423 |        "kid11_15_Y                     17241.0  0.090308  0.286631  0.0  0.000000   \n",
3424 |        "kid11_15_nan                   17241.0  0.017748  0.132039  0.0  0.000000   \n",
3425 |        "kid16_17_U                     17241.0  0.886085  0.317717  0.0  1.000000   \n",
3426 |        "kid16_17_Y                     17241.0  0.096166  0.294827  0.0  0.000000   \n",
3427 |        "kid16_17_nan                   17241.0  0.017748  0.132039  0.0  0.000000   \n",
3428 |        "car_buy_New                    17241.0  0.422017  0.493896  0.0  0.000000   \n",
3429 |        "car_buy_UNKNOWN                17241.0  0.560234  0.496373  0.0  0.000000   \n",
3430 |        "car_buy_nan                    17241.0  0.017748  0.132039  0.0  0.000000   \n",
3431 |        "\n",
3432 |        "                                    50%       75%  max  \n",
3433 |        "LOCAL_CUR_SAV_SLOPE            0.333333  0.358169  1.0  \n",
3434 |        "LOCAL_CUR_MON_AVG_BAL          0.000425  0.003037  1.0  \n",
3435 |        "LOCAL_OVEONEYR_FF_MON_AVG_BAL  0.008991  0.017082  1.0  \n",
3436 |        "LOCAL_FIX_MON_AVG_BAL          0.001300  0.002320  1.0  \n",
3437 |        "LOCAL_FIX_MON_AVG_BAL_PROP     0.854396  0.990436  1.0  \n",
3438 |        "LOCAL_BELONEYR_FF_SLOPE        0.344437  0.344437  1.0  \n",
3439 |        "LOCAL_BELONEYR_FF_MON_AVG_BAL  0.000000  0.000000  1.0  \n",
3440 |        "LOCAL_OVEONEYR_FF_SLOPE        0.344442  0.345026  1.0  \n",
3441 |        "LOCAL_SAV_SLOPE                0.344535  0.356539  1.0  \n",
3442 |        "LOCAL_SAV_CUR_ALL_BAL          0.001710  0.002807  1.0  \n",
3443 |        "LOCAL_SAV_MON_AVG_BAL          0.001749  0.002861  1.0  \n",
3444 |        "SAV_SLOPE                      0.344535  0.356519  1.0  \n",
3445 |        "SAV_CUR_ALL_BAL                0.001713  0.002809  1.0  \n",
3446 |        "SAV_MON_AVG_BAL                0.001751  0.002862  1.0  \n",
3447 |        "FR_SAV_CUR_ALL_BAL             0.000000  0.000000  1.0  \n",
3448 |        "ASSET_CUR_ALL_BAL              0.001781  0.002941  1.0  \n",
3449 |        "ASSET_MON_AVG_BAL              0.001816  0.002988  1.0  \n",
3450 |        "LOCAL_CUR_TRANS_TX_AMT         0.000000  0.000000  1.0  \n",
3451 |        "LOCAL_CUR_LASTSAV_TX_AMT       0.000000  0.000000  1.0  \n",
3452 |        "LOCAL_CUR_WITHDRAW_TX_AMT      0.000000  0.000492  1.0  \n",
3453 |        "LOCAL_FIX_OPEN_ACC_TX_AMT      0.000000  0.000000  1.0  \n",
3454 |        "LOCAL_FIX_WITHDRAW_TX_AMT      0.000000  0.000000  1.0  \n",
3455 |        "LOCAL_FIX_CLOSE_ACC_TX_AMT     0.000000  0.000000  1.0  \n",
3456 |        "L6M_INDFINA_ALL_TX_AMT         0.000000  0.000000  1.0  \n",
3457 |        "da_Mean                        0.004292  0.017167  1.0  \n",
3458 |        "ovrmou_Mean                    0.001457  0.021592  1.0  \n",
3459 |        "ovrrev_Mean                    0.001622  0.022766  1.0  \n",
3460 |        "vceovr_Mean                    0.000873  0.021976  1.0  \n",
3461 |        "datovr_Mean                    0.000000  0.000000  1.0  \n",
3462 |        "roam_Mean                      0.000000  0.000619  1.0  \n",
3463 |        "...                                 ...       ...  ...  \n",
3464 |        "marital_nan                    0.000000  0.000000  1.0  \n",
3465 |        "wrkwoman_Y                     0.000000  0.000000  1.0  \n",
3466 |        "wrkwoman_nan                   1.000000  1.000000  1.0  \n",
3467 |        "proptype_A                     0.000000  1.000000  1.0  \n",
3468 |        "proptype_B                     0.000000  0.000000  1.0  \n",
3469 |        "proptype_D                     0.000000  0.000000  1.0  \n",
3470 |        "proptype_E                     0.000000  0.000000  1.0  \n",
3471 |        "proptype_G                     0.000000  0.000000  1.0  \n",
3472 |        "proptype_M                     0.000000  0.000000  1.0  \n",
3473 |        "proptype_nan                   1.000000  1.000000  1.0  \n",
3474 |        "pcowner_Y                      0.000000  0.000000  1.0  \n",
3475 |        "pcowner_nan                    1.000000  1.000000  1.0  \n",
3476 |        "kid0_2_U                       1.000000  1.000000  1.0  \n",
3477 |        "kid0_2_Y                       0.000000  0.000000  1.0  \n",
3478 |        "kid0_2_nan                     0.000000  0.000000  1.0  \n",
3479 |        "kid3_5_U                       1.000000  1.000000  1.0  \n",
3480 |        "kid3_5_Y                       0.000000  0.000000  1.0  \n",
3481 |        "kid3_5_nan                     0.000000  0.000000  1.0  \n",
3482 |        "kid6_10_U                      1.000000  1.000000  1.0  \n",
3483 |        "kid6_10_Y                      0.000000  0.000000  1.0  \n",
3484 |        "kid6_10_nan                    0.000000  0.000000  1.0  \n",
3485 |        "kid11_15_U                     1.000000  1.000000  1.0  \n",
3486 |        "kid11_15_Y                     0.000000  0.000000  1.0  \n",
3487 |        "kid11_15_nan                   0.000000  0.000000  1.0  \n",
3488 |        "kid16_17_U                     1.000000  1.000000  1.0  \n",
3489 |        "kid16_17_Y                     0.000000  0.000000  1.0  \n",
3490 |        "kid16_17_nan                   0.000000  0.000000  1.0  \n",
3491 |        "car_buy_New                    0.000000  1.000000  1.0  \n",
3492 |        "car_buy_UNKNOWN                1.000000  1.000000  1.0  \n",
3493 |        "car_buy_nan                    0.000000  0.000000  1.0  \n",
3494 |        "\n",
3495 |        "[178 rows x 8 columns]"
3496 |       ]
3497 |      },
3498 |      "execution_count": 27,
3499 |      "metadata": {},
3500 |      "output_type": "execute_result"
3501 |     }
3502 |    ],
3503 |    "source": [
3504 |     "model_data_des"
3505 |    ]
3506 |   },
3507 |   {
3508 |    "cell_type": "code",
3509 |    "execution_count": null,
3510 |    "metadata": {
3511 |     "collapsed": true
3512 |    },
3513 |    "outputs": [],
3514 |    "source": []
3515 |   },
3516 |   {
3517 |    "cell_type": "code",
3518 |    "execution_count": null,
3519 |    "metadata": {
3520 |     "collapsed": true
3521 |    },
3522 |    "outputs": [],
3523 |    "source": []
3524 |   }
3525 |  ],
3526 |  "metadata": {
3527 |   "kernelspec": {
3528 |    "display_name": "Python 3",
3529 |    "language": "python",
3530 |    "name": "python3"
3531 |   },
3532 |   "language_info": {
3533 |    "codemirror_mode": {
3534 |     "name": "ipython",
3535 |     "version": 3
3536 |    },
3537 |    "file_extension": ".py",
3538 |    "mimetype": "text/x-python",
3539 |    "name": "python",
3540 |    "nbconvert_exporter": "python",
3541 |    "pygments_lexer": "ipython3",
3542 |    "version": "3.6.3"
3543 |   }
3544 |  },
3545 |  "nbformat": 4,
3546 |  "nbformat_minor": 2
3547 | }
3548 | 


--------------------------------------------------------------------------------