notebooks/Robustness_Study_HelocFico.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "2089a6db-afb4-46e4-ba3a-a185a95842c5",
   "metadata": {},
   "source": [
    "# Robustness Study: Heloc Fico dataset\n",
    "For this dataset, the target variable to predict is a binary variable called RiskPerformance. The value “Bad” indicates that a consumer was 90 days past due or worse at least once over a period of 24 months from when the credit account was opened. The value “Good” indicates that they have made their payments without ever being more than 90 days overdue. https://community.fico.com/s/explainable-machine-learning-challenge."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2987b76c-72f3-4f6d-9702-fdea7d7c42b7",
   "metadata": {},
   "source": [
    "## Librairies "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "82d04152-4cfa-4340-9324-53731fffe8b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os, sys, time\n",
    "import numpy as np\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "3aa685ff-4be5-4319-9469-4bb59e5efddc",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import roc_auc_score, average_precision_score\n",
    "from sklearn.preprocessing import StandardScaler, OneHotEncoder\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "from sklearn import set_config\n",
    "from sklearn.compose import ColumnTransformer\n",
    "from sklearn.pipeline import make_pipeline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "66ecf9e5-af68-49a9-92ec-dc6f423b4ac7",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch \n",
    "import torch.nn as nn\n",
    "import torch.optim as optim "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "4683e029-2e4c-4021-b582-c669f242de76",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "import torch\n",
    "from torch.optim.lr_scheduler import ReduceLROnPlateau\n",
    "from skorch.callbacks import EarlyStopping,LRScheduler,Checkpoint, TrainEndCheckpoint, EpochScoring\n",
    "from skorch.dataset import Dataset\n",
    "from skorch.helper import predefined_split\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import mean_squared_error, r2_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "8fb63796-266f-48d2-94d0-ba2b4e525f90",
   "metadata": {},
   "outputs": [],
   "source": [
    "sys.path.append(\"../\")\n",
    "from sramodels.SRAModels import TabSRALinearClassifier,InputShapeSetterTabSRA,LinearClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "a2b1616d-6738-4447-8422-23cb43c176d2",
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.options.display.max_rows = 1000\n",
    "pd.options.display.max_columns = 1000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "8eda4b5a-4dbf-4a8f-925a-b377bc223556",
   "metadata": {},
   "outputs": [],
   "source": [
    "from interpret.glassbox import ExplainableBoostingClassifier, ExplainableBoostingRegressor\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "78e160eb-487c-4e2b-8264-a98c28db7c27",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/kodjo/anaconda3/lib/python3.9/site-packages/xgboost/compat.py:36: FutureWarning: pandas.Int64Index is deprecated and will be removed from pandas in a future version. Use pandas.Index with the appropriate dtype instead.\n",
      "  from pandas import MultiIndex, Int64Index\n"
     ]
    }
   ],
   "source": [
    "from xgboost import XGBClassifier\n",
    "import xgboost as xgb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "e72ce230-24ea-45de-b9b4-e1255afbaa16",
   "metadata": {},
   "outputs": [],
   "source": [
    "import shap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "0dfe0958-a5ab-4660-8cbc-3bbece20e4e9",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "plt.style.use('ggplot')\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c9d6b675-c0f6-4822-8411-9b5abe80e4c3",
   "metadata": {},
   "source": [
    "## Data loading, preprocessing and Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "cbb697a4-ef56-4f4a-bcf3-5672157d20a9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>RiskPerformance</th>\n",
       "      <th>ExternalRiskEstimate</th>\n",
       "      <th>MSinceOldestTradeOpen</th>\n",
       "      <th>MSinceMostRecentTradeOpen</th>\n",
       "      <th>AverageMInFile</th>\n",
       "      <th>NumSatisfactoryTrades</th>\n",
       "      <th>NumTrades60Ever2DerogPubRec</th>\n",
       "      <th>NumTrades90Ever2DerogPubRec</th>\n",
       "      <th>PercentTradesNeverDelq</th>\n",
       "      <th>MSinceMostRecentDelq</th>\n",
       "      <th>MaxDelq2PublicRecLast12M</th>\n",
       "      <th>MaxDelqEver</th>\n",
       "      <th>NumTotalTrades</th>\n",
       "      <th>NumTradesOpeninLast12M</th>\n",
       "      <th>PercentInstallTrades</th>\n",
       "      <th>MSinceMostRecentInqexcl7days</th>\n",
       "      <th>NumInqLast6M</th>\n",
       "      <th>NumInqLast6Mexcl7days</th>\n",
       "      <th>NetFractionRevolvingBurden</th>\n",
       "      <th>NetFractionInstallBurden</th>\n",
       "      <th>NumRevolvingTradesWBalance</th>\n",
       "      <th>NumInstallTradesWBalance</th>\n",
       "      <th>NumBank2NatlTradesWHighUtilization</th>\n",
       "      <th>PercentTradesWBalance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>55</td>\n",
       "      <td>144</td>\n",
       "      <td>4</td>\n",
       "      <td>84</td>\n",
       "      <td>20</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>83</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>23</td>\n",
       "      <td>1</td>\n",
       "      <td>43</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>33</td>\n",
       "      <td>-8</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>61</td>\n",
       "      <td>58</td>\n",
       "      <td>15</td>\n",
       "      <td>41</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>100</td>\n",
       "      <td>-7</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>67</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-8</td>\n",
       "      <td>0</td>\n",
       "      <td>-8</td>\n",
       "      <td>-8</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>67</td>\n",
       "      <td>66</td>\n",
       "      <td>5</td>\n",
       "      <td>24</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>100</td>\n",
       "      <td>-7</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>4</td>\n",
       "      <td>44</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>53</td>\n",
       "      <td>66</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>66</td>\n",
       "      <td>169</td>\n",
       "      <td>1</td>\n",
       "      <td>73</td>\n",
       "      <td>28</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>93</td>\n",
       "      <td>76</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "      <td>30</td>\n",
       "      <td>3</td>\n",
       "      <td>57</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>72</td>\n",
       "      <td>83</td>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>91</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>81</td>\n",
       "      <td>333</td>\n",
       "      <td>27</td>\n",
       "      <td>132</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>100</td>\n",
       "      <td>-7</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>25</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>51</td>\n",
       "      <td>89</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>80</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   RiskPerformance  ExternalRiskEstimate  MSinceOldestTradeOpen  \\\n",
       "0                0                    55                    144   \n",
       "1                0                    61                     58   \n",
       "2                0                    67                     66   \n",
       "3                0                    66                    169   \n",
       "4                0                    81                    333   \n",
       "\n",
       "   MSinceMostRecentTradeOpen  AverageMInFile  NumSatisfactoryTrades  \\\n",
       "0                          4              84                     20   \n",
       "1                         15              41                      2   \n",
       "2                          5              24                      9   \n",
       "3                          1              73                     28   \n",
       "4                         27             132                     12   \n",
       "\n",
       "   NumTrades60Ever2DerogPubRec  NumTrades90Ever2DerogPubRec  \\\n",
       "0                            3                            0   \n",
       "1                            4                            4   \n",
       "2                            0                            0   \n",
       "3                            1                            1   \n",
       "4                            0                            0   \n",
       "\n",
       "   PercentTradesNeverDelq  MSinceMostRecentDelq  MaxDelq2PublicRecLast12M  \\\n",
       "0                      83                     2                         3   \n",
       "1                     100                    -7                         0   \n",
       "2                     100                    -7                         7   \n",
       "3                      93                    76                         6   \n",
       "4                     100                    -7                         7   \n",
       "\n",
       "   MaxDelqEver  NumTotalTrades  NumTradesOpeninLast12M  PercentInstallTrades  \\\n",
       "0            5              23                       1                    43   \n",
       "1            8               7                       0                    67   \n",
       "2            8               9                       4                    44   \n",
       "3            6              30                       3                    57   \n",
       "4            8              12                       0                    25   \n",
       "\n",
       "   MSinceMostRecentInqexcl7days  NumInqLast6M  NumInqLast6Mexcl7days  \\\n",
       "0                             0             0                      0   \n",
       "1                             0             0                      0   \n",
       "2                             0             4                      4   \n",
       "3                             0             5                      4   \n",
       "4                             0             1                      1   \n",
       "\n",
       "   NetFractionRevolvingBurden  NetFractionInstallBurden  \\\n",
       "0                          33                        -8   \n",
       "1                           0                        -8   \n",
       "2                          53                        66   \n",
       "3                          72                        83   \n",
       "4                          51                        89   \n",
       "\n",
       "   NumRevolvingTradesWBalance  NumInstallTradesWBalance  \\\n",
       "0                           8                         1   \n",
       "1                           0                        -8   \n",
       "2                           4                         2   \n",
       "3                           6                         4   \n",
       "4                           3                         1   \n",
       "\n",
       "   NumBank2NatlTradesWHighUtilization  PercentTradesWBalance  \n",
       "0                                   1                     69  \n",
       "1                                  -8                      0  \n",
       "2                                   1                     86  \n",
       "3                                   3                     91  \n",
       "4                                   0                     80  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_path = '../ESANN_XKDD/Datasets/'\n",
    "data = pd.read_csv(data_path+'heloc_dataset_v1.csv')\n",
    "data[\"RiskPerformance\"] =  np.where(data[\"RiskPerformance\"]=='Bad',0,1)\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "de235b18-5e74-45e2-ac00-4b25e5495d47",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.4780571756382063"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.RiskPerformance.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "20281c88-7cef-4213-8120-665ed4cb4b92",
   "metadata": {},
   "source": [
    "### Preprocessing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "94b1d691-2026-4691-a5c2-60bb73b785cc",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_copy = data.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "7226af8b-7e49-48ff-8067-de351d18fe2c",
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_names = list(data.columns)\n",
    "feature_names.remove('RiskPerformance')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8a9a935b-105f-459f-9d7b-a94f090461ec",
   "metadata": {},
   "source": [
    "### Split "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "d5fbb86b-3ad8-4293-972b-c08dbdcccc24",
   "metadata": {},
   "outputs": [],
   "source": [
    "k =5\n",
    "Sf= StratifiedKFold(n_splits=k, shuffle=True, random_state= 42)\n",
    "#datasets = [(X[train_index], X[test_index],y[train_index], y[test_index]) for train_index,test_index in Sf.split(X, y) ]\n",
    "datasets_raw = [(data.iloc[train_index], data.iloc[test_index]) for train_index,test_index in Sf.split(data, data.RiskPerformance.values) ]\n",
    "data_train_df,data_test_df = datasets_raw[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "fa27a6d4-0a06-4d40-be5c-74ebf519f089",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_train_df,data_test_df = datasets_raw[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "3d76fc9f-467b-4bad-8684-27f509e12e68",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_, X_test_, Y_train_, Y_test_ = data_train_df[feature_names], data_test_df[feature_names],data_train_df.RiskPerformance.values,data_test_df.RiskPerformance.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "1591692a-74f7-4ac3-8e08-ec14bb672a36",
   "metadata": {},
   "outputs": [],
   "source": [
    "set_config(transform_output=\"pandas\")\n",
    "processor =  StandardScaler()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "9a3d7f5e-6855-4c72-8092-4db7c3f56537",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_ = processor.fit_transform(X_train_)\n",
    "X_test_ = processor.transform(X_test_)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "60720d5d-f257-4b56-9eb8-56af62522a38",
   "metadata": {},
   "source": [
    "### Correlation visualization (map)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "ee94046c-3fcf-433e-a38e-7adafd26086b",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "f13aa89c-f8cc-4f8a-9a73-9fa287fb91be",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_tmp = X_train_.copy()\n",
    "X_train_tmp['y'] = Y_train_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "b2e00f21-16eb-4102-80d3-d5b675b6144d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_13896/1798339133.py:6: DeprecationWarning: `np.bool` is a deprecated alias for the builtin `bool`. To silence this warning, use `bool` by itself. Doing this will not modify any behavior and is safe. If you specifically wanted the numpy scalar type, use `np.bool_` here.\n",
      "Deprecated in NumPy 1.20; for more details and guidance: https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations\n",
      "  sns.heatmap(corr, mask=np.zeros_like(corr, dtype=np.bool), cmap=sns.diverging_palette(220, 10, as_cmap=True),\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "\n",
    "f, ax = plt.subplots(figsize=(12, 12))\n",
    "tmp = X_train_tmp.astype(float) \n",
    "corr = tmp.corr(method='spearman')\n",
    "sns.heatmap(corr, mask=np.zeros_like(corr, dtype=np.bool), cmap=sns.diverging_palette(220, 10, as_cmap=True),\n",
    "            square=True, ax=ax);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "5f626d5c-2780-479c-8870-0aecc2452c79",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_13896/2226804073.py:6: DeprecationWarning: `np.bool` is a deprecated alias for the builtin `bool`. To silence this warning, use `bool` by itself. Doing this will not modify any behavior and is safe. If you specifically wanted the numpy scalar type, use `np.bool_` here.\n",
      "Deprecated in NumPy 1.20; for more details and guidance: https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations\n",
      "  sns.heatmap(corr, mask=np.zeros_like(corr, dtype=np.bool), cmap=sns.diverging_palette(220, 10, as_cmap=True),\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "\n",
    "f, ax = plt.subplots(figsize=(12, 12))\n",
    "tmp = X_train_tmp.astype(float) \n",
    "corr = tmp.corr()\n",
    "sns.heatmap(corr, mask=np.zeros_like(corr, dtype=np.bool), cmap=sns.diverging_palette(220, 10, as_cmap=True),\n",
    "            square=True, ax=ax);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "accepting-mexico",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c5357995-647b-493b-b60e-b350a463e33f",
   "metadata": {},
   "source": [
    "### Linear  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "90551aab-c952-4220-9133-f771899996dc",
   "metadata": {},
   "outputs": [],
   "source": [
    "#valid_loss\n",
    "other_params_linear = {\"module__bias\":True,\n",
    "                \"optimizer__lr\":0.007143256650213327,\n",
    "                \"optimizer__weight_decay\":5.348330773994543e-06,\n",
    "                \"random_state\":42,\n",
    "                \"criterion\": nn.BCEWithLogitsLoss,\n",
    "                \"module__negative_indice\":None,\n",
    "                \"module__positive_indice\":None,\n",
    "                \"max_epochs\":500,\n",
    "                \"batch_size\":256,\n",
    "                \"device\":'cuda'#change the device in 'cpu' if there is no GPU\n",
    "               }\n",
    "scoring = EpochScoring(scoring='roc_auc',lower_is_better=False)\n",
    "setter = InputShapeSetterTabSRA(regression=False)\n",
    "early_stop = EarlyStopping(monitor=scoring.scoring, patience=50,load_best=True,lower_is_better=False, threshold=0.0001,threshold_mode='abs')\n",
    "lr_scheduler = LRScheduler(policy=ReduceLROnPlateau, patience=25, min_lr=2e-5,factor=0.2, verbose=1)\n",
    "call_b = [scoring, setter, early_stop, lr_scheduler]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "8bd4929b-0bb2-4d03-a39f-ee37270dace8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2092,)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y_test_.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "126a3369-ddde-4559-8e15-f21530b4673a",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Re-initializing module because the following parameters were re-set: module__dim_input, module__dim_output.\n",
      "Re-initializing criterion.\n",
      "Re-initializing optimizer.\n",
      "  epoch    roc_auc    train_loss    valid_acc    valid_loss     dur\n",
      "-------  ---------  ------------  -----------  ------------  ------\n",
      "      1     \u001b[36m0.7447\u001b[0m        \u001b[32m0.6349\u001b[0m       \u001b[35m0.6840\u001b[0m        \u001b[31m0.5956\u001b[0m  0.6451\n",
      "      2     \u001b[36m0.7680\u001b[0m        \u001b[32m0.5829\u001b[0m       \u001b[35m0.7012\u001b[0m        \u001b[31m0.5772\u001b[0m  0.0590\n",
      "      3     \u001b[36m0.7725\u001b[0m        \u001b[32m0.5744\u001b[0m       \u001b[35m0.7094\u001b[0m        \u001b[31m0.5720\u001b[0m  0.0583\n",
      "      4     \u001b[36m0.7750\u001b[0m        \u001b[32m0.5703\u001b[0m       \u001b[35m0.7098\u001b[0m        \u001b[31m0.5696\u001b[0m  0.0554\n",
      "      5     \u001b[36m0.7771\u001b[0m        \u001b[32m0.5686\u001b[0m       \u001b[35m0.7122\u001b[0m        \u001b[31m0.5680\u001b[0m  0.0595\n",
      "      6     \u001b[36m0.7773\u001b[0m        \u001b[32m0.5671\u001b[0m       0.7118        \u001b[31m0.5675\u001b[0m  0.0600\n",
      "      7     \u001b[36m0.7776\u001b[0m        0.5671       0.7113        \u001b[31m0.5670\u001b[0m  0.0607\n",
      "      8     \u001b[36m0.7792\u001b[0m        0.5677       \u001b[35m0.7137\u001b[0m        \u001b[31m0.5661\u001b[0m  0.0602\n",
      "      9     0.7768        \u001b[32m0.5659\u001b[0m       0.7103        0.5674  0.0597\n",
      "     10     0.7791        \u001b[32m0.5653\u001b[0m       0.7127        0.5662  0.0580\n",
      "     11     0.7754        \u001b[32m0.5652\u001b[0m       0.7065        0.5690  0.0628\n",
      "     12     0.7791        0.5657       0.7113        \u001b[31m0.5659\u001b[0m  0.0808\n",
      "     13     0.7786        \u001b[32m0.5650\u001b[0m       0.7118        0.5665  0.0574\n",
      "     14     0.7737        0.5655       0.7075        0.5705  0.0537\n",
      "     15     \u001b[36m0.7796\u001b[0m        \u001b[32m0.5645\u001b[0m       0.7094        0.5661  0.0505\n",
      "     16     \u001b[36m0.7804\u001b[0m        \u001b[32m0.5643\u001b[0m       0.7094        \u001b[31m0.5659\u001b[0m  0.0572\n",
      "     17     0.7800        \u001b[32m0.5643\u001b[0m       0.7122        \u001b[31m0.5654\u001b[0m  0.0620\n",
      "     18     0.7801        0.5646       0.7132        0.5659  0.0719\n",
      "     19     0.7798        0.5647       0.7103        0.5656  0.0504\n",
      "     20     0.7797        \u001b[32m0.5638\u001b[0m       0.7084        0.5657  0.0557\n",
      "     21     0.7797        0.5639       0.7118        0.5659  0.0646\n",
      "     22     0.7801        0.5642       0.7098        \u001b[31m0.5654\u001b[0m  0.0588\n",
      "     23     0.7774        \u001b[32m0.5637\u001b[0m       0.7084        0.5674  0.0676\n",
      "     24     0.7803        0.5652       0.7113        0.5656  0.0634\n",
      "     25     0.7798        0.5640       0.7118        0.5657  0.0698\n",
      "     26     0.7800        0.5647       0.7108        0.5655  0.0625\n",
      "     27     0.7791        \u001b[32m0.5637\u001b[0m       0.7113        0.5665  0.0652\n",
      "     28     0.7792        0.5639       0.7118        0.5663  0.0644\n",
      "     29     0.7787        0.5638       0.7098        0.5669  0.0787\n",
      "     30     0.7802        0.5640       0.7127        0.5656  0.0625\n",
      "     31     0.7803        0.5639       0.7118        0.5657  0.0618\n",
      "     32     0.7798        0.5638       0.7098        0.5657  0.0571\n",
      "     33     \u001b[36m0.7805\u001b[0m        \u001b[32m0.5634\u001b[0m       0.7137        0.5657  0.0608\n",
      "     34     0.7796        0.5636       0.7118        0.5661  0.0571\n",
      "     35     0.7801        0.5640       0.7103        0.5657  0.0587\n",
      "     36     0.7797        0.5638       0.7094        0.5664  0.0566\n",
      "     37     0.7802        0.5637       0.7113        0.5657  0.0583\n",
      "     38     0.7792        0.5635       0.7122        0.5668  0.0606\n",
      "     39     \u001b[36m0.7809\u001b[0m        0.5635       0.7127        0.5656  0.0604\n",
      "     40     \u001b[36m0.7811\u001b[0m        0.5637       0.7122        0.5656  0.0595\n",
      "     41     0.7801        0.5640       0.7118        0.5658  0.0612\n",
      "     42     0.7808        0.5640       \u001b[35m0.7146\u001b[0m        0.5655  0.0614\n",
      "     43     0.7793        0.5645       0.7113        0.5662  0.0642\n",
      "     44     0.7801        0.5637       0.7122        0.5656  0.0748\n",
      "     45     0.7806        0.5634       \u001b[35m0.7151\u001b[0m        0.5655  0.0606\n",
      "     46     0.7808        0.5634       0.7108        0.5658  0.0670\n",
      "     47     0.7808        0.5639       0.7127        0.5655  0.0631\n",
      "     48     0.7795        0.5635       0.7094        0.5664  0.0703\n",
      "     49     0.7761        0.5636       0.7094        0.5685  0.0617\n",
      "     50     0.7805        0.5649       0.7122        0.5657  0.0673\n",
      "     51     0.7797        0.5635       0.7113        0.5663  0.0686\n",
      "     52     0.7792        0.5635       0.7098        0.5666  0.0670\n",
      "     53     0.7809        0.5637       0.7132        0.5655  0.0612\n",
      "     54     0.7776        0.5636       0.7108        0.5678  0.0609\n",
      "     55     0.7811        0.5638       \u001b[35m0.7156\u001b[0m        0.5655  0.0617\n",
      "     56     0.7796        0.5636       0.7098        0.5664  0.0681\n",
      "     57     0.7807        0.5645       0.7151        0.5656  0.0667\n",
      "     58     0.7802        0.5641       0.7122        0.5659  0.0579\n",
      "Epoch 00059: reducing learning rate of group 0 to 1.4287e-03.\n",
      "     59     0.7807        0.5635       0.7122        0.5658  0.0618\n",
      "     60     0.7798        0.5634       0.7098        0.5661  0.0626\n",
      "     61     0.7804        \u001b[32m0.5629\u001b[0m       0.7118        0.5656  0.0582\n",
      "     62     0.7802        \u001b[32m0.5628\u001b[0m       0.7137        0.5657  0.0601\n",
      "     63     0.7809        0.5628       0.7141        \u001b[31m0.5654\u001b[0m  0.0739\n",
      "     64     0.7802        0.5628       0.7113        0.5659  0.0846\n",
      "     65     0.7807        0.5628       0.7127        0.5655  0.0663\n",
      "     66     0.7802        \u001b[32m0.5627\u001b[0m       0.7132        0.5657  0.0637\n",
      "     67     0.7807        0.5628       0.7127        0.5655  0.0612\n",
      "     68     0.7804        \u001b[32m0.5627\u001b[0m       0.7118        0.5656  0.0606\n",
      "     69     0.7804        0.5629       0.7127        0.5656  0.0617\n",
      "     70     0.7805        0.5627       0.7132        0.5656  0.0595\n",
      "     71     0.7808        0.5628       0.7141        0.5655  0.0586\n",
      "     72     0.7803        0.5627       0.7118        0.5657  0.0613\n",
      "     73     0.7808        0.5627       0.7127        0.5655  0.0638\n",
      "     74     0.7805        0.5629       0.7132        0.5655  0.0585\n",
      "     75     0.7809        \u001b[32m0.5627\u001b[0m       0.7137        0.5655  0.0601\n",
      "     76     0.7800        0.5628       0.7113        0.5658  0.0626\n",
      "     77     0.7800        0.5629       0.7118        0.5660  0.0705\n",
      "     78     0.7810        0.5628       0.7141        0.5654  0.0892\n",
      "     79     0.7803        0.5627       0.7113        0.5657  0.0744\n",
      "     80     0.7809        0.5630       0.7127        0.5655  0.0723\n",
      "     81     0.7801        0.5633       0.7122        0.5658  0.0681\n",
      "     82     0.7809        \u001b[32m0.5626\u001b[0m       0.7146        0.5654  0.0695\n",
      "     83     0.7809        0.5628       0.7137        0.5655  0.0639\n",
      "     84     0.7800        0.5627       0.7118        0.5660  0.0580\n",
      "     85     0.7809        0.5629       0.7137        0.5655  0.0553\n",
      "     86     0.7804        0.5628       0.7127        0.5657  0.0650\n",
      "     87     0.7811        0.5627       0.7146        \u001b[31m0.5654\u001b[0m  0.0654\n",
      "     88     0.7805        0.5627       0.7151        0.5655  0.0680\n",
      "     89     0.7803        0.5628       0.7122        0.5657  0.0599\n",
      "Stopping since roc_auc has not improved in the last 50 epochs.\n",
      "Restoring best model from epoch 40.\n",
      "CPU times: user 6.58 s, sys: 426 ms, total: 7.01 s\n",
      "Wall time: 7.45 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "valid_dataset = Dataset(X_test_.values.astype(np.float32),Y_test_.astype(np.float32))\n",
    "LR = LinearClassifier(**other_params_linear,train_split = predefined_split(valid_dataset),callbacks = call_b)#,\n",
    "_ = LR.fit(X_train_.values.astype(np.float32),Y_train_.astype(np.float32))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "d3120b1b-e7ec-47f8-a7b7-b845be5cd917",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "23"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(feature_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "8f29bc22-d680-45c1-9eb4-5bed2868c85e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ExternalRiskEstimate</th>\n",
       "      <th>MSinceOldestTradeOpen</th>\n",
       "      <th>MSinceMostRecentTradeOpen</th>\n",
       "      <th>AverageMInFile</th>\n",
       "      <th>NumSatisfactoryTrades</th>\n",
       "      <th>NumTrades60Ever2DerogPubRec</th>\n",
       "      <th>NumTrades90Ever2DerogPubRec</th>\n",
       "      <th>PercentTradesNeverDelq</th>\n",
       "      <th>MSinceMostRecentDelq</th>\n",
       "      <th>MaxDelq2PublicRecLast12M</th>\n",
       "      <th>MaxDelqEver</th>\n",
       "      <th>NumTotalTrades</th>\n",
       "      <th>NumTradesOpeninLast12M</th>\n",
       "      <th>PercentInstallTrades</th>\n",
       "      <th>MSinceMostRecentInqexcl7days</th>\n",
       "      <th>NumInqLast6M</th>\n",
       "      <th>NumInqLast6Mexcl7days</th>\n",
       "      <th>NetFractionRevolvingBurden</th>\n",
       "      <th>NetFractionInstallBurden</th>\n",
       "      <th>NumRevolvingTradesWBalance</th>\n",
       "      <th>NumInstallTradesWBalance</th>\n",
       "      <th>NumBank2NatlTradesWHighUtilization</th>\n",
       "      <th>PercentTradesWBalance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.810069</td>\n",
       "      <td>0.057554</td>\n",
       "      <td>-0.134708</td>\n",
       "      <td>0.345282</td>\n",
       "      <td>0.370131</td>\n",
       "      <td>-0.319619</td>\n",
       "      <td>-0.280662</td>\n",
       "      <td>0.259453</td>\n",
       "      <td>-0.02527</td>\n",
       "      <td>0.24565</td>\n",
       "      <td>-0.24878</td>\n",
       "      <td>-0.027017</td>\n",
       "      <td>-0.089957</td>\n",
       "      <td>-0.260288</td>\n",
       "      <td>0.13948</td>\n",
       "      <td>-0.330407</td>\n",
       "      <td>-0.085133</td>\n",
       "      <td>-0.416432</td>\n",
       "      <td>-0.026475</td>\n",
       "      <td>-0.41397</td>\n",
       "      <td>0.08272</td>\n",
       "      <td>0.095416</td>\n",
       "      <td>-0.004345</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   ExternalRiskEstimate  MSinceOldestTradeOpen  MSinceMostRecentTradeOpen  \\\n",
       "0              0.810069               0.057554                  -0.134708   \n",
       "\n",
       "   AverageMInFile  NumSatisfactoryTrades  NumTrades60Ever2DerogPubRec  \\\n",
       "0        0.345282               0.370131                    -0.319619   \n",
       "\n",
       "   NumTrades90Ever2DerogPubRec  PercentTradesNeverDelq  MSinceMostRecentDelq  \\\n",
       "0                    -0.280662                0.259453              -0.02527   \n",
       "\n",
       "   MaxDelq2PublicRecLast12M  MaxDelqEver  NumTotalTrades  \\\n",
       "0                   0.24565     -0.24878       -0.027017   \n",
       "\n",
       "   NumTradesOpeninLast12M  PercentInstallTrades  MSinceMostRecentInqexcl7days  \\\n",
       "0               -0.089957             -0.260288                       0.13948   \n",
       "\n",
       "   NumInqLast6M  NumInqLast6Mexcl7days  NetFractionRevolvingBurden  \\\n",
       "0     -0.330407              -0.085133                   -0.416432   \n",
       "\n",
       "   NetFractionInstallBurden  NumRevolvingTradesWBalance  \\\n",
       "0                 -0.026475                    -0.41397   \n",
       "\n",
       "   NumInstallTradesWBalance  NumBank2NatlTradesWHighUtilization  \\\n",
       "0                   0.08272                            0.095416   \n",
       "\n",
       "   PercentTradesWBalance  \n",
       "0              -0.004345  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Coef  = pd.DataFrame(LR.get_weights()[0])\n",
    "Coef = Coef\n",
    "Coef.columns=feature_names\n",
    "Coef"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "800cac39-0b66-4f3b-9f17-a4545ab5649b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--AUCROC perf = 0.7811135531135531 -- AUCPR perf = 0.7571985643576569\n"
     ]
    }
   ],
   "source": [
    "# Inference on the test set. the Predict function will return the ouptut scores and the attention weights as Model.get_attention is True\n",
    "pred_lr = LR.predict_proba(X_test_.values.astype(np.float32))[:,1]\n",
    "\n",
    "best_aucroc_lr = roc_auc_score(Y_test_, pred_lr) \n",
    "best_aurpr_lr = average_precision_score(Y_test_, pred_lr) \n",
    "print(f\"--AUCROC perf = {best_aucroc_lr} -- AUCPR perf = {best_aurpr_lr}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b0c40273-23e2-4ca1-ad51-58bf966a7891",
   "metadata": {},
   "source": [
    "### TabSRA  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "439db264-3d29-4739-9df4-53c42c311afe",
   "metadata": {},
   "outputs": [],
   "source": [
    "#valid_loss\n",
    "other_params = {\"module__encoder_bias\":True,\n",
    "                \"module__classifier_bias\":True,\n",
    "                \"module__n_head\":1,\n",
    "                \"module__dim_head\":8,\n",
    "                \"module__n_hidden_encoder\":1,\n",
    "                \"module__dropout_rate\":0.3,\n",
    "                \"optimizer__lr\":0.009,#275745003824123,\n",
    "                \"optimizer__weight_decay\":0.0004,#338871767939995,\n",
    "                \"random_state\":42,\n",
    "                \"criterion\": nn.BCEWithLogitsLoss,\n",
    "                \"module__negative_indice\":None,\n",
    "                \"module__positive_indice\":None,\n",
    "                \"max_epochs\":500,\n",
    "                \"batch_size\":1792,\n",
    "                \"device\":'cuda'#change the device in 'cpu' if there is no GPU\n",
    "               }\n",
    "scoring = EpochScoring(scoring='roc_auc',lower_is_better=False)\n",
    "setter = InputShapeSetterTabSRA(regression=False)\n",
    "early_stop = EarlyStopping(monitor=scoring.scoring, patience=50,load_best=True,lower_is_better=False, threshold=0.0001,threshold_mode='abs')\n",
    "lr_scheduler = LRScheduler(policy=ReduceLROnPlateau, patience=25, min_lr=2e-5,factor=0.2, verbose=1)\n",
    "call_b = [scoring, setter, early_stop, lr_scheduler]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "85b4aac5-a7d4-4ce8-9200-4b29b221d403",
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Re-initializing module because the following parameters were re-set: module__dim_input, module__dim_output.\n",
      "Re-initializing criterion.\n",
      "Re-initializing optimizer.\n",
      "  epoch    roc_auc    train_loss    valid_acc    valid_loss     dur\n",
      "-------  ---------  ------------  -----------  ------------  ------\n",
      "      1     \u001b[36m0.6811\u001b[0m        \u001b[32m0.6932\u001b[0m       \u001b[35m0.6233\u001b[0m        \u001b[31m0.6731\u001b[0m  0.0614\n",
      "      2     \u001b[36m0.7341\u001b[0m        \u001b[32m0.6594\u001b[0m       \u001b[35m0.6721\u001b[0m        \u001b[31m0.6391\u001b[0m  0.0559\n",
      "      3     \u001b[36m0.7431\u001b[0m        \u001b[32m0.6271\u001b[0m       \u001b[35m0.6821\u001b[0m        \u001b[31m0.6138\u001b[0m  0.0492\n",
      "      4     \u001b[36m0.7539\u001b[0m        \u001b[32m0.6065\u001b[0m       \u001b[35m0.6898\u001b[0m        \u001b[31m0.5996\u001b[0m  0.0493\n",
      "      5     \u001b[36m0.7625\u001b[0m        \u001b[32m0.5945\u001b[0m       \u001b[35m0.7012\u001b[0m        \u001b[31m0.5902\u001b[0m  0.0527\n",
      "      6     \u001b[36m0.7706\u001b[0m        \u001b[32m0.5860\u001b[0m       \u001b[35m0.7051\u001b[0m        \u001b[31m0.5824\u001b[0m  0.0487\n",
      "      7     \u001b[36m0.7742\u001b[0m        \u001b[32m0.5789\u001b[0m       \u001b[35m0.7089\u001b[0m        \u001b[31m0.5756\u001b[0m  0.0462\n",
      "      8     \u001b[36m0.7784\u001b[0m        \u001b[32m0.5723\u001b[0m       \u001b[35m0.7141\u001b[0m        \u001b[31m0.5700\u001b[0m  0.0524\n",
      "      9     \u001b[36m0.7806\u001b[0m        \u001b[32m0.5663\u001b[0m       \u001b[35m0.7165\u001b[0m        \u001b[31m0.5661\u001b[0m  0.0470\n",
      "     10     \u001b[36m0.7825\u001b[0m        \u001b[32m0.5627\u001b[0m       0.7156        \u001b[31m0.5637\u001b[0m  0.1106\n",
      "     11     \u001b[36m0.7851\u001b[0m        \u001b[32m0.5601\u001b[0m       0.7165        \u001b[31m0.5613\u001b[0m  0.0439\n",
      "     12     \u001b[36m0.7861\u001b[0m        \u001b[32m0.5575\u001b[0m       \u001b[35m0.7189\u001b[0m        \u001b[31m0.5591\u001b[0m  0.0447\n",
      "     13     \u001b[36m0.7877\u001b[0m        \u001b[32m0.5562\u001b[0m       0.7189        \u001b[31m0.5576\u001b[0m  0.0493\n",
      "     14     0.7859        \u001b[32m0.5545\u001b[0m       \u001b[35m0.7208\u001b[0m        0.5579  0.0709\n",
      "     15     \u001b[36m0.7885\u001b[0m        0.5549       0.7175        \u001b[31m0.5559\u001b[0m  0.0471\n",
      "     16     0.7877        \u001b[32m0.5528\u001b[0m       \u001b[35m0.7228\u001b[0m        0.5561  0.0487\n",
      "     17     0.7881        0.5535       \u001b[35m0.7242\u001b[0m        \u001b[31m0.5558\u001b[0m  0.0480\n",
      "     18     \u001b[36m0.7901\u001b[0m        \u001b[32m0.5525\u001b[0m       0.7199        \u001b[31m0.5544\u001b[0m  0.0510\n",
      "     19     0.7895        \u001b[32m0.5516\u001b[0m       0.7204        0.5545  0.0477\n",
      "     20     0.7898        \u001b[32m0.5506\u001b[0m       0.7199        0.5544  0.0493\n",
      "     21     0.7891        \u001b[32m0.5505\u001b[0m       0.7208        \u001b[31m0.5535\u001b[0m  0.0507\n",
      "     22     0.7899        \u001b[32m0.5495\u001b[0m       0.7237        \u001b[31m0.5532\u001b[0m  0.0488\n",
      "     23     \u001b[36m0.7902\u001b[0m        0.5496       \u001b[35m0.7266\u001b[0m        \u001b[31m0.5530\u001b[0m  0.0491\n",
      "     24     \u001b[36m0.7903\u001b[0m        \u001b[32m0.5493\u001b[0m       0.7247        \u001b[31m0.5529\u001b[0m  0.0458\n",
      "     25     \u001b[36m0.7911\u001b[0m        \u001b[32m0.5490\u001b[0m       0.7228        \u001b[31m0.5518\u001b[0m  0.0503\n",
      "     26     0.7909        \u001b[32m0.5477\u001b[0m       0.7256        0.5522  0.0464\n",
      "     27     0.7909        \u001b[32m0.5477\u001b[0m       0.7247        0.5520  0.0494\n",
      "     28     0.7903        0.5482       0.7223        0.5524  0.0477\n",
      "     29     0.7905        0.5487       0.7251        0.5522  0.0497\n",
      "     30     \u001b[36m0.7916\u001b[0m        0.5486       0.7251        \u001b[31m0.5516\u001b[0m  0.0482\n",
      "     31     0.7912        0.5479       0.7223        \u001b[31m0.5515\u001b[0m  0.0494\n",
      "     32     0.7910        \u001b[32m0.5464\u001b[0m       0.7237        0.5516  0.0498\n",
      "     33     \u001b[36m0.7916\u001b[0m        0.5476       0.7247        0.5522  0.0445\n",
      "     34     0.7905        0.5468       0.7251        0.5519  0.0497\n",
      "     35     \u001b[36m0.7920\u001b[0m        0.5467       0.7251        \u001b[31m0.5507\u001b[0m  0.0440\n",
      "     36     0.7911        0.5472       0.7237        0.5518  0.0483\n",
      "     37     0.7913        0.5469       0.7213        0.5509  0.0476\n",
      "     38     0.7919        0.5466       0.7223        0.5514  0.0525\n",
      "     39     0.7918        \u001b[32m0.5458\u001b[0m       0.7242        0.5513  0.0522\n",
      "     40     0.7909        0.5467       0.7251        0.5517  0.1090\n",
      "     41     0.7917        \u001b[32m0.5454\u001b[0m       0.7218        0.5515  0.0488\n",
      "     42     0.7911        0.5465       0.7204        0.5513  0.0476\n",
      "     43     0.7918        0.5465       0.7218        \u001b[31m0.5507\u001b[0m  0.0494\n",
      "     44     0.7909        0.5467       0.7208        0.5521  0.0503\n",
      "     45     \u001b[36m0.7921\u001b[0m        0.5467       0.7223        0.5512  0.0474\n",
      "     46     0.7916        0.5465       0.7213        0.5511  0.0430\n",
      "     47     0.7913        \u001b[32m0.5450\u001b[0m       0.7218        0.5513  0.0507\n",
      "     48     0.7915        0.5456       0.7218        0.5508  0.0613\n",
      "     49     0.7919        0.5453       0.7213        0.5509  0.0522\n",
      "     50     0.7919        0.5465       0.7208        0.5511  0.0419\n",
      "     51     0.7915        0.5456       0.7194        0.5515  0.0481\n",
      "     52     \u001b[36m0.7922\u001b[0m        0.5459       0.7228        \u001b[31m0.5504\u001b[0m  0.0407\n",
      "     53     0.7919        0.5460       0.7185        0.5517  0.0475\n",
      "     54     \u001b[36m0.7922\u001b[0m        0.5453       0.7199        0.5511  0.0456\n",
      "     55     0.7917        \u001b[32m0.5450\u001b[0m       0.7223        0.5512  0.0463\n",
      "     56     \u001b[36m0.7924\u001b[0m        \u001b[32m0.5449\u001b[0m       0.7208        0.5504  0.0482\n",
      "     57     0.7914        \u001b[32m0.5440\u001b[0m       0.7232        0.5512  0.0479\n",
      "     58     0.7924        0.5446       0.7194        0.5508  0.0461\n",
      "     59     0.7919        0.5454       0.7228        0.5506  0.0445\n",
      "     60     0.7906        0.5447       0.7218        0.5520  0.0443\n",
      "     61     0.7922        0.5448       0.7213        0.5506  0.0444\n",
      "     62     0.7918        0.5449       0.7213        0.5512  0.0456\n",
      "     63     \u001b[36m0.7929\u001b[0m        0.5452       0.7204        \u001b[31m0.5497\u001b[0m  0.0642\n",
      "     64     0.7907        0.5447       0.7199        0.5520  0.0479\n",
      "     65     0.7926        0.5453       0.7232        0.5504  0.0476\n",
      "     66     0.7923        \u001b[32m0.5435\u001b[0m       0.7213        0.5507  0.0456\n",
      "     67     0.7920        0.5448       0.7185        0.5510  0.0440\n",
      "     68     0.7916        \u001b[32m0.5433\u001b[0m       0.7208        0.5511  0.0422\n",
      "     69     0.7917        0.5444       0.7175        0.5512  0.0466\n",
      "     70     0.7920        0.5455       0.7213        0.5509  0.0465\n",
      "     71     \u001b[36m0.7930\u001b[0m        0.5447       0.7218        0.5506  0.0470\n",
      "     72     0.7920        0.5454       0.7194        0.5509  0.0413\n",
      "     73     0.7916        0.5443       0.7208        0.5507  0.0492\n",
      "     74     0.7921        0.5443       0.7194        0.5507  0.0465\n",
      "     75     0.7917        0.5447       0.7213        0.5513  0.0591\n",
      "     76     0.7917        0.5443       0.7180        0.5515  0.0550\n",
      "     77     0.7922        0.5447       0.7223        0.5506  0.0500\n",
      "     78     0.7920        0.5449       0.7185        0.5508  0.0496\n",
      "     79     0.7921        0.5438       0.7180        0.5509  0.0471\n",
      "     80     0.7929        0.5440       0.7180        0.5502  0.0517\n",
      "     81     0.7914        0.5446       0.7208        0.5511  0.0479\n",
      "     82     0.7917        0.5437       0.7165        0.5517  0.0507\n",
      "     83     0.7918        0.5438       0.7189        0.5509  0.0519\n",
      "     84     \u001b[36m0.7930\u001b[0m        \u001b[32m0.5428\u001b[0m       0.7208        0.5503  0.0477\n",
      "     85     0.7919        0.5436       0.7170        0.5513  0.0641\n",
      "     86     0.7928        0.5435       0.7204        0.5506  0.0453\n",
      "     87     0.7925        0.5431       0.7213        0.5506  0.0541\n",
      "     88     0.7919        0.5437       0.7218        0.5508  0.0529\n",
      "     89     0.7921        \u001b[32m0.5426\u001b[0m       0.7185        0.5509  0.0553\n",
      "     90     0.7922        0.5430       0.7199        0.5511  0.0526\n",
      "     91     0.7922        0.5445       0.7213        0.5507  0.0475\n",
      "     92     0.7920        \u001b[32m0.5420\u001b[0m       0.7189        0.5516  0.0503\n",
      "     93     0.7923        0.5437       0.7213        0.5506  0.0546\n",
      "     94     0.7912        0.5432       0.7185        0.5514  0.0666\n",
      "     95     0.7926        0.5428       0.7194        0.5503  0.0648\n",
      "     96     0.7928        0.5422       0.7194        0.5503  0.1122\n",
      "     97     0.7922        0.5434       0.7232        0.5507  0.0488\n",
      "     98     \u001b[36m0.7931\u001b[0m        0.5423       0.7199        0.5499  0.0494\n",
      "     99     0.7928        0.5429       0.7189        0.5504  0.0476\n",
      "    100     0.7920        0.5425       0.7208        0.5512  0.0500\n",
      "    101     0.7925        0.5429       0.7180        0.5506  0.0571\n",
      "    102     0.7922        \u001b[32m0.5415\u001b[0m       0.7199        0.5507  0.0745\n",
      "    103     0.7909        0.5426       0.7180        0.5517  0.0569\n",
      "    104     0.7925        0.5430       0.7213        0.5502  0.0573\n",
      "    105     0.7919        0.5415       0.7208        0.5510  0.0752\n",
      "    106     0.7918        0.5431       0.7189        0.5510  0.0573\n",
      "    107     \u001b[36m0.7931\u001b[0m        0.5427       0.7199        0.5503  0.0568\n",
      "    108     0.7924        0.5436       0.7194        0.5505  0.0582\n",
      "    109     0.7920        0.5416       0.7189        0.5517  0.0781\n",
      "    110     0.7930        0.5437       0.7213        0.5502  0.0568\n",
      "    111     0.7928        0.5423       0.7175        0.5505  0.0503\n",
      "    112     0.7908        0.5431       0.7194        0.5516  0.0534\n",
      "    113     0.7927        0.5419       0.7189        0.5511  0.0539\n",
      "    114     0.7925        0.5435       0.7204        0.5502  0.0567\n",
      "    115     0.7925        0.5424       0.7185        0.5512  0.0572\n",
      "    116     0.7925        \u001b[32m0.5411\u001b[0m       0.7199        0.5504  0.0712\n",
      "    117     0.7923        0.5418       0.7199        0.5506  0.0496\n",
      "    118     0.7924        0.5419       0.7208        0.5505  0.0474\n",
      "    119     0.7919        0.5423       0.7156        0.5511  0.0487\n",
      "    120     0.7915        0.5418       0.7208        0.5515  0.0505\n",
      "    121     0.7927        0.5413       0.7185        0.5504  0.0502\n",
      "    122     0.7924        0.5421       0.7180        0.5512  0.0817\n",
      "    123     0.7912        0.5420       0.7218        0.5515  0.0750\n",
      "    124     0.7921        0.5416       0.7161        0.5514  0.0487\n",
      "    125     0.7921        0.5426       0.7204        0.5507  0.0483\n",
      "    126     0.7930        0.5414       0.7199        0.5498  0.1119\n",
      "    127     0.7917        0.5413       0.7199        0.5509  0.0474\n",
      "    128     0.7931        0.5425       0.7208        \u001b[31m0.5497\u001b[0m  0.0552\n",
      "    129     0.7926        0.5416       0.7208        0.5505  0.0692\n",
      "    130     0.7926        0.5413       0.7189        0.5508  0.0577\n",
      "    131     0.7925        \u001b[32m0.5407\u001b[0m       0.7208        0.5505  0.0536\n",
      "    132     0.7926        0.5411       0.7180        0.5506  0.0652\n",
      "    133     0.7907        0.5420       0.7204        0.5522  0.0566\n",
      "    134     0.7927        0.5414       0.7208        0.5500  0.0544\n",
      "    135     0.7903        0.5409       0.7218        0.5530  0.0619\n",
      "    136     0.7923        0.5429       0.7175        0.5509  0.0658\n",
      "    137     0.7897        0.5413       0.7208        0.5527  0.0528\n",
      "    138     0.7924        \u001b[32m0.5405\u001b[0m       0.7185        0.5509  0.0490\n",
      "    139     0.7913        0.5416       0.7199        0.5512  0.0530\n",
      "    140     \u001b[36m0.7932\u001b[0m        0.5414       0.7170        0.5501  0.0475\n",
      "    141     0.7916        0.5412       0.7218        0.5514  0.0591\n",
      "    142     \u001b[36m0.7933\u001b[0m        0.5412       0.7213        \u001b[31m0.5494\u001b[0m  0.0524\n",
      "    143     0.7927        0.5410       0.7213        0.5503  0.0489\n",
      "    144     0.7923        \u001b[32m0.5403\u001b[0m       0.7218        0.5508  0.0453\n",
      "    145     0.7923        0.5420       0.7204        0.5506  0.0507\n",
      "    146     0.7924        \u001b[32m0.5399\u001b[0m       0.7199        0.5510  0.0600\n",
      "    147     \u001b[36m0.7936\u001b[0m        0.5405       0.7218        0.5496  0.0680\n",
      "    148     0.7933        0.5401       0.7223        0.5499  0.0456\n",
      "    149     0.7931        0.5408       0.7223        0.5500  0.0504\n",
      "    150     0.7907        0.5417       0.7199        0.5523  0.0507\n",
      "    151     0.7929        0.5415       0.7232        0.5504  0.0481\n",
      "    152     0.7916        0.5403       0.7175        0.5513  0.0439\n",
      "    153     0.7934        0.5405       0.7194        0.5498  0.0564\n",
      "    154     0.7933        0.5400       0.7218        0.5497  0.0504\n",
      "    155     0.7927        0.5401       0.7199        0.5503  0.0466\n",
      "    156     0.7930        0.5403       0.7237        0.5499  0.0450\n",
      "    157     0.7929        0.5402       0.7180        0.5500  0.0449\n",
      "    158     0.7935        0.5400       0.7199        0.5498  0.0562\n",
      "    159     0.7933        0.5410       0.7204        0.5496  0.0561\n",
      "    160     \u001b[36m0.7939\u001b[0m        0.5406       0.7218        \u001b[31m0.5488\u001b[0m  0.0543\n",
      "    161     0.7928        0.5402       0.7218        0.5506  0.0516\n",
      "    162     0.7930        0.5402       0.7204        0.5500  0.0505\n",
      "    163     0.7938        0.5413       0.7204        0.5493  0.0484\n",
      "    164     0.7913        0.5422       0.7223        0.5513  0.0681\n",
      "    165     0.7935        0.5406       0.7194        0.5498  0.0542\n",
      "    166     0.7931        0.5400       0.7208        0.5498  0.0540\n",
      "    167     0.7936        0.5409       0.7223        0.5493  0.0520\n",
      "    168     0.7934        0.5405       0.7218        0.5495  0.0712\n",
      "    169     0.7930        \u001b[32m0.5396\u001b[0m       0.7213        0.5500  0.0535\n",
      "    170     0.7919        0.5399       0.7242        0.5508  0.0558\n",
      "    171     0.7932        0.5413       0.7185        0.5506  0.0533\n",
      "    172     0.7925        0.5403       0.7194        0.5509  0.0681\n",
      "    173     \u001b[36m0.7939\u001b[0m        0.5404       0.7218        0.5498  0.0520\n",
      "    174     0.7931        0.5405       0.7223        0.5501  0.0550\n",
      "    175     0.7928        0.5407       0.7194        0.5500  0.0534\n",
      "    176     0.7934        0.5401       0.7194        0.5503  0.0542\n",
      "    177     0.7932        0.5407       0.7218        0.5497  0.0636\n",
      "    178     0.7934        0.5410       0.7185        0.5499  0.0483\n",
      "    179     \u001b[36m0.7940\u001b[0m        0.5399       0.7237        0.5490  0.0468\n",
      "    180     0.7918        \u001b[32m0.5393\u001b[0m       0.7204        0.5507  0.0620\n",
      "    181     0.7936        0.5406       0.7218        0.5494  0.0576\n",
      "    182     0.7933        0.5398       0.7228        0.5498  0.1295\n",
      "    183     0.7933        0.5397       0.7213        0.5500  0.0492\n",
      "    184     0.7935        0.5398       0.7218        0.5499  0.0516\n",
      "    185     0.7932        0.5401       0.7232        0.5496  0.0507\n",
      "    186     0.7931        0.5399       0.7228        0.5500  0.0523\n",
      "    187     \u001b[36m0.7941\u001b[0m        0.5400       0.7218        0.5491  0.0672\n",
      "    188     0.7927        0.5404       0.7194        0.5506  0.0523\n",
      "    189     0.7936        0.5400       0.7237        0.5497  0.0544\n",
      "    190     0.7930        0.5400       0.7213        0.5501  0.0500\n",
      "    191     0.7935        \u001b[32m0.5391\u001b[0m       0.7213        0.5497  0.0626\n",
      "    192     0.7929        0.5394       0.7218        0.5495  0.0542\n",
      "    193     0.7934        0.5398       0.7237        0.5494  0.0558\n",
      "    194     \u001b[36m0.7944\u001b[0m        0.5393       0.7247        \u001b[31m0.5487\u001b[0m  0.0513\n",
      "    195     0.7938        0.5393       0.7228        0.5492  0.0628\n",
      "    196     0.7942        \u001b[32m0.5385\u001b[0m       0.7237        \u001b[31m0.5486\u001b[0m  0.0545\n",
      "    197     0.7936        0.5394       0.7213        0.5488  0.0496\n",
      "    198     0.7927        0.5394       0.7228        0.5500  0.0511\n",
      "    199     0.7936        0.5391       0.7232        0.5494  0.0619\n",
      "    200     0.7937        0.5399       0.7228        0.5492  0.0546\n",
      "    201     0.7936        0.5398       0.7213        0.5494  0.0557\n",
      "    202     0.7938        0.5394       0.7208        0.5496  0.0519\n",
      "    203     0.7925        0.5390       0.7232        0.5502  0.0587\n",
      "    204     0.7936        0.5399       0.7204        0.5496  0.0473\n",
      "    205     0.7934        0.5398       0.7213        0.5492  0.0459\n",
      "    206     0.7938        0.5389       0.7223        0.5491  0.0541\n",
      "    207     0.7940        \u001b[32m0.5384\u001b[0m       0.7242        0.5491  0.0556\n",
      "    208     0.7939        0.5400       0.7228        0.5492  0.0573\n",
      "    209     0.7943        0.5392       0.7213        0.5489  0.0611\n",
      "    210     0.7934        0.5394       0.7199        0.5495  0.0617\n",
      "    211     0.7934        \u001b[32m0.5381\u001b[0m       0.7204        0.5493  0.0531\n",
      "    212     0.7935        0.5386       0.7204        0.5494  0.1139\n",
      "    213     0.7939        0.5393       0.7237        0.5496  0.0584\n",
      "    214     0.7940        0.5386       0.7251        0.5492  0.0531\n",
      "    215     0.7933        0.5387       0.7228        0.5497  0.0533\n",
      "    216     0.7923        0.5386       0.7232        0.5509  0.0552\n",
      "    217     0.7933        \u001b[32m0.5378\u001b[0m       0.7189        0.5501  0.0626\n",
      "    218     0.7915        0.5387       0.7204        0.5511  0.0538\n",
      "    219     \u001b[36m0.7947\u001b[0m        0.5385       0.7218        \u001b[31m0.5486\u001b[0m  0.0624\n",
      "    220     0.7944        0.5389       0.7218        0.5491  0.0714\n",
      "    221     0.7927        0.5394       0.7208        0.5505  0.0651\n",
      "    222     0.7938        0.5390       0.7237        0.5493  0.0536\n",
      "    223     0.7934        0.5393       0.7213        0.5501  0.0513\n",
      "    224     0.7940        0.5384       0.7228        0.5491  0.0542\n",
      "    225     0.7925        0.5388       0.7189        0.5508  0.0531\n",
      "    226     0.7938        0.5395       0.7213        0.5493  0.0518\n",
      "    227     0.7945        0.5396       0.7232        0.5489  0.0489\n",
      "    228     0.7929        0.5379       0.7189        0.5502  0.0551\n",
      "    229     0.7940        0.5383       0.7237        0.5487  0.0584\n",
      "    230     0.7943        0.5384       0.7228        0.5488  0.0712\n",
      "    231     0.7939        0.5383       0.7232        0.5490  0.0612\n",
      "    232     0.7935        0.5385       0.7218        0.5495  0.0559\n",
      "    233     0.7942        \u001b[32m0.5376\u001b[0m       0.7223        0.5493  0.0549\n",
      "    234     0.7939        \u001b[32m0.5375\u001b[0m       0.7213        0.5492  0.0686\n",
      "    235     0.7936        0.5379       0.7208        0.5497  0.0537\n",
      "    236     0.7940        0.5379       0.7199        0.5490  0.0539\n",
      "    237     0.7940        0.5379       0.7208        0.5490  0.0520\n",
      "    238     0.7939        0.5377       0.7223        0.5492  0.0694\n",
      "    239     0.7932        0.5378       0.7213        0.5495  0.0508\n",
      "    240     0.7929        0.5378       0.7213        0.5501  0.0485\n",
      "    241     0.7942        0.5386       0.7237        0.5487  0.0608\n",
      "    242     0.7942        0.5378       0.7232        0.5491  0.0643\n",
      "    243     0.7941        \u001b[32m0.5371\u001b[0m       0.7237        0.5490  0.0506\n",
      "    244     0.7946        0.5372       0.7232        \u001b[31m0.5483\u001b[0m  0.0501\n",
      "    245     0.7940        0.5374       0.7223        0.5487  0.0667\n",
      "    246     0.7936        0.5378       0.7213        0.5493  0.0561\n",
      "    247     0.7946        0.5374       0.7218        0.5489  0.0505\n",
      "    248     0.7939        \u001b[32m0.5364\u001b[0m       0.7223        0.5491  0.0526\n",
      "    249     0.7928        0.5375       0.7204        0.5503  0.0626\n",
      "    250     0.7944        0.5366       0.7185        0.5490  0.0584\n",
      "    251     0.7936        0.5378       0.7204        0.5495  0.0503\n",
      "    252     0.7936        0.5373       0.7218        0.5493  0.0513\n",
      "    253     0.7930        0.5383       0.7228        0.5500  0.0641\n",
      "    254     0.7937        0.5370       0.7204        0.5499  0.0565\n",
      "    255     0.7936        0.5378       0.7208        0.5491  0.0532\n",
      "    256     0.7943        0.5379       0.7237        0.5489  0.0516\n",
      "    257     0.7937        0.5372       0.7213        0.5492  0.0561\n",
      "    258     0.7945        0.5382       0.7223        0.5490  0.0611\n",
      "    259     0.7935        0.5366       0.7208        0.5495  0.0529\n",
      "    260     0.7933        0.5378       0.7189        0.5500  0.0494\n",
      "    261     0.7946        0.5374       0.7208        0.5488  0.0598\n",
      "    262     0.7938        0.5366       0.7223        0.5493  0.0539\n",
      "    263     0.7941        0.5382       0.7213        0.5497  0.0510\n",
      "    264     0.7929        0.5369       0.7161        0.5504  0.0523\n",
      "    265     0.7933        0.5374       0.7194        0.5494  0.0654\n",
      "    266     0.7943        0.5373       0.7180        0.5490  0.0530\n",
      "    267     0.7935        0.5372       0.7199        0.5491  0.0513\n",
      "    268     0.7936        0.5379       0.7232        0.5493  0.1267\n",
      "Stopping since roc_auc has not improved in the last 50 epochs.\n",
      "Restoring best model from epoch 219.\n",
      "CPU times: user 15.7 s, sys: 73.6 ms, total: 15.8 s\n",
      "Wall time: 15.8 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "valid_dataset = Dataset(X_test_.values.astype(np.float32),Y_test_.astype(np.float32))\n",
    "TabSRA = TabSRALinearClassifier(**other_params,train_split = predefined_split(valid_dataset),callbacks = call_b)#,\n",
    "_ = TabSRA.fit(X_train_.values.astype(np.float32),Y_train_.astype(np.float32))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "05237092-3cb3-4403-a0f9-f3271c329678",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--AUCROC perf = 0.7947014652014652 -- AUCPR perf = 0.7753445622806545\n"
     ]
    }
   ],
   "source": [
    "# Inference on the test set. the Predict function will return the ouptut scores and the attention weights as Model.get_attention is True\n",
    "pred_tabsra = TabSRA.predict_proba(X_test_.values.astype(np.float32))[:,1]\n",
    "\n",
    "best_aucroc = roc_auc_score(Y_test_, pred_tabsra) \n",
    "best_aurpr = average_precision_score(Y_test_, pred_tabsra) \n",
    "print(f\"--AUCROC perf = {best_aucroc} -- AUCPR perf = {best_aurpr}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "3a7e9d03-8f51-444d-bc72-891832eaf6e3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ExternalRiskEstimate</th>\n",
       "      <th>MSinceOldestTradeOpen</th>\n",
       "      <th>MSinceMostRecentTradeOpen</th>\n",
       "      <th>AverageMInFile</th>\n",
       "      <th>NumSatisfactoryTrades</th>\n",
       "      <th>NumTrades60Ever2DerogPubRec</th>\n",
       "      <th>NumTrades90Ever2DerogPubRec</th>\n",
       "      <th>PercentTradesNeverDelq</th>\n",
       "      <th>MSinceMostRecentDelq</th>\n",
       "      <th>MaxDelq2PublicRecLast12M</th>\n",
       "      <th>MaxDelqEver</th>\n",
       "      <th>NumTotalTrades</th>\n",
       "      <th>NumTradesOpeninLast12M</th>\n",
       "      <th>PercentInstallTrades</th>\n",
       "      <th>MSinceMostRecentInqexcl7days</th>\n",
       "      <th>NumInqLast6M</th>\n",
       "      <th>NumInqLast6Mexcl7days</th>\n",
       "      <th>NetFractionRevolvingBurden</th>\n",
       "      <th>NetFractionInstallBurden</th>\n",
       "      <th>NumRevolvingTradesWBalance</th>\n",
       "      <th>NumInstallTradesWBalance</th>\n",
       "      <th>NumBank2NatlTradesWHighUtilization</th>\n",
       "      <th>PercentTradesWBalance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.611713</td>\n",
       "      <td>0.144753</td>\n",
       "      <td>-0.2851</td>\n",
       "      <td>1.447543</td>\n",
       "      <td>0.918009</td>\n",
       "      <td>-0.210805</td>\n",
       "      <td>-0.08536</td>\n",
       "      <td>1.79208</td>\n",
       "      <td>0.396699</td>\n",
       "      <td>0.863246</td>\n",
       "      <td>-0.022624</td>\n",
       "      <td>-0.061059</td>\n",
       "      <td>0.152821</td>\n",
       "      <td>-0.725277</td>\n",
       "      <td>0.896041</td>\n",
       "      <td>-0.773791</td>\n",
       "      <td>-0.218277</td>\n",
       "      <td>-1.111185</td>\n",
       "      <td>-0.157643</td>\n",
       "      <td>-1.151793</td>\n",
       "      <td>0.170946</td>\n",
       "      <td>0.030289</td>\n",
       "      <td>0.057035</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   ExternalRiskEstimate  MSinceOldestTradeOpen  MSinceMostRecentTradeOpen  \\\n",
       "0              1.611713               0.144753                    -0.2851   \n",
       "\n",
       "   AverageMInFile  NumSatisfactoryTrades  NumTrades60Ever2DerogPubRec  \\\n",
       "0        1.447543               0.918009                    -0.210805   \n",
       "\n",
       "   NumTrades90Ever2DerogPubRec  PercentTradesNeverDelq  MSinceMostRecentDelq  \\\n",
       "0                     -0.08536                 1.79208              0.396699   \n",
       "\n",
       "   MaxDelq2PublicRecLast12M  MaxDelqEver  NumTotalTrades  \\\n",
       "0                  0.863246    -0.022624       -0.061059   \n",
       "\n",
       "   NumTradesOpeninLast12M  PercentInstallTrades  MSinceMostRecentInqexcl7days  \\\n",
       "0                0.152821             -0.725277                      0.896041   \n",
       "\n",
       "   NumInqLast6M  NumInqLast6Mexcl7days  NetFractionRevolvingBurden  \\\n",
       "0     -0.773791              -0.218277                   -1.111185   \n",
       "\n",
       "   NetFractionInstallBurden  NumRevolvingTradesWBalance  \\\n",
       "0                 -0.157643                   -1.151793   \n",
       "\n",
       "   NumInstallTradesWBalance  NumBank2NatlTradesWHighUtilization  \\\n",
       "0                  0.170946                            0.030289   \n",
       "\n",
       "   PercentTradesWBalance  \n",
       "0               0.057035  "
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Coef  = pd.DataFrame(TabSRA.get_weights()[0])\n",
    "Coef = Coef.T\n",
    "Coef.columns=feature_names\n",
    "Coef"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "72af0eb1-646b-4f57-9deb-6cc304bf5f8d",
   "metadata": {},
   "source": [
    "### XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "a2272e98-4ce4-4869-a759-49cbb1cb2530",
   "metadata": {},
   "outputs": [],
   "source": [
    "static_params = {\n",
    "                 'random_state':42,\n",
    "                 'objective': 'binary:logistic',\n",
    "                 'n_jobs': -1,\n",
    "                 #'base_score':y.mean()\n",
    "\n",
    "}\n",
    "\n",
    "params = {'eta': 0.0054836062607363095, 'max_depth': 4, 'colsample_bytree': 1.0, 'subsample': 0.8, 'n_estimators': 1100, 'min_child_weight': 22}\n",
    "\n",
    "\n",
    "\n",
    "params.update(static_params)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "037dc1ca-95f0-4d4b-8e9b-56c3a031962a",
   "metadata": {},
   "outputs": [],
   "source": [
    "model_xgb = XGBClassifier(**params)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "07b84507-14b3-4e2b-85f9-67b14c4214c6",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/kodjo/anaconda3/lib/python3.9/site-packages/xgboost/sklearn.py:1224: UserWarning: The use of label encoder in XGBClassifier is deprecated and will be removed in a future release. To remove this warning, do the following: 1) Pass option use_label_encoder=False when constructing XGBClassifier object; and 2) Encode your labels (y) as integers starting with 0, i.e. 0, 1, 2, ..., [num_class - 1].\n",
      "  warnings.warn(label_encoder_deprecation_msg, UserWarning)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-2 {color: black;background-color: white;}#sk-container-id-2 pre{padding: 0;}#sk-container-id-2 div.sk-toggleable {background-color: white;}#sk-container-id-2 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-2 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-2 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-2 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-2 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-2 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-2 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-2 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-2 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-2 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-2 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-2 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-2 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-2 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-2 div.sk-item {position: relative;z-index: 1;}#sk-container-id-2 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-2 div.sk-item::before, #sk-container-id-2 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-2 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-2 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-2 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-2 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-2 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-2 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-2 div.sk-label-container {text-align: center;}#sk-container-id-2 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-2 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-2\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>XGBClassifier(base_score=0.5, booster=&#x27;gbtree&#x27;, colsample_bylevel=1,\n",
       "              colsample_bynode=1, colsample_bytree=1.0,\n",
       "              enable_categorical=False, eta=0.0054836062607363095, gamma=0,\n",
       "              gpu_id=-1, importance_type=None, interaction_constraints=&#x27;&#x27;,\n",
       "              learning_rate=0.00548360636, max_delta_step=0, max_depth=4,\n",
       "              min_child_weight=22, missing=nan, monotone_constraints=&#x27;()&#x27;,\n",
       "              n_estimators=1100, n_jobs=-1, num_parallel_tree=1,\n",
       "              predictor=&#x27;auto&#x27;, random_state=42, reg_alpha=0, reg_lambda=1,\n",
       "              scale_pos_weight=1, subsample=0.8, tree_method=&#x27;exact&#x27;,\n",
       "              validate_parameters=1, verbosity=None)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" checked><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">XGBClassifier</label><div class=\"sk-toggleable__content\"><pre>XGBClassifier(base_score=0.5, booster=&#x27;gbtree&#x27;, colsample_bylevel=1,\n",
       "              colsample_bynode=1, colsample_bytree=1.0,\n",
       "              enable_categorical=False, eta=0.0054836062607363095, gamma=0,\n",
       "              gpu_id=-1, importance_type=None, interaction_constraints=&#x27;&#x27;,\n",
       "              learning_rate=0.00548360636, max_delta_step=0, max_depth=4,\n",
       "              min_child_weight=22, missing=nan, monotone_constraints=&#x27;()&#x27;,\n",
       "              n_estimators=1100, n_jobs=-1, num_parallel_tree=1,\n",
       "              predictor=&#x27;auto&#x27;, random_state=42, reg_alpha=0, reg_lambda=1,\n",
       "              scale_pos_weight=1, subsample=0.8, tree_method=&#x27;exact&#x27;,\n",
       "              validate_parameters=1, verbosity=None)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "XGBClassifier(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
       "              colsample_bynode=1, colsample_bytree=1.0,\n",
       "              enable_categorical=False, eta=0.0054836062607363095, gamma=0,\n",
       "              gpu_id=-1, importance_type=None, interaction_constraints='',\n",
       "              learning_rate=0.00548360636, max_delta_step=0, max_depth=4,\n",
       "              min_child_weight=22, missing=nan, monotone_constraints='()',\n",
       "              n_estimators=1100, n_jobs=-1, num_parallel_tree=1,\n",
       "              predictor='auto', random_state=42, reg_alpha=0, reg_lambda=1,\n",
       "              scale_pos_weight=1, subsample=0.8, tree_method='exact',\n",
       "              validate_parameters=1, verbosity=None)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "eval_set = [(X_train_.values, Y_train_), (X_test_.values, Y_test_)] #(X_train, Y_train[:,1]), \n",
    "model_xgb.fit(X_train_.values, Y_train_, eval_metric=['auc'], eval_set=eval_set, verbose=0,early_stopping_rounds=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "cc9eafe4-73b1-4310-98c8-f82333f9ab7a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--AUCROC perf = 0.7976245421245421 -- AUCPR perf = 0.7814848170562338\n"
     ]
    }
   ],
   "source": [
    "pred_xgb = model_xgb.predict_proba(X_test_)[:,1]\n",
    "best_aucroc_xgb = roc_auc_score(Y_test_, pred_xgb) \n",
    "best_aurpr_xgb = average_precision_score(Y_test_, pred_xgb) \n",
    "print(f\"--AUCROC perf = {best_aucroc_xgb} -- AUCPR perf = {best_aurpr_xgb}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5368c205-95c6-4b6e-ad41-a5c8722f61bb",
   "metadata": {},
   "source": [
    "### EBM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "9275b5ab-2046-4f66-b4a3-64cca009171a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-3 {color: black;background-color: white;}#sk-container-id-3 pre{padding: 0;}#sk-container-id-3 div.sk-toggleable {background-color: white;}#sk-container-id-3 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-3 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-3 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-3 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-3 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-3 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-3 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-3 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-3 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-3 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-3 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-3 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-3 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-3 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-3 div.sk-item {position: relative;z-index: 1;}#sk-container-id-3 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-3 div.sk-item::before, #sk-container-id-3 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-3 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-3 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-3 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-3 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-3 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-3 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-3 div.sk-label-container {text-align: center;}#sk-container-id-3 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-3 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-3\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>ExplainableBoostingClassifier(inner_bags=25, max_rounds=20000, n_jobs=-1,\n",
       "                              outer_bags=100)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" checked><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ExplainableBoostingClassifier</label><div class=\"sk-toggleable__content\"><pre>ExplainableBoostingClassifier(inner_bags=25, max_rounds=20000, n_jobs=-1,\n",
       "                              outer_bags=100)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "ExplainableBoostingClassifier(inner_bags=25, max_rounds=20000, n_jobs=-1,\n",
       "                              outer_bags=100)"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ExplainableBoostingClassifier(interactions=10,max_rounds=20000,early_stopping_rounds=50,inner_bags=25,outer_bags=100,n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "10518a11-8b27-4d46-b4f6-141415a23ba4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 309 ms, sys: 320 ms, total: 628 ms\n",
      "Wall time: 22.9 s\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-4 {color: black;background-color: white;}#sk-container-id-4 pre{padding: 0;}#sk-container-id-4 div.sk-toggleable {background-color: white;}#sk-container-id-4 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-4 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-4 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-4 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-4 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-4 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-4 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-4 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-4 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-4 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-4 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-4 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-4 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-4 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-4 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-4 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-4 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-4 div.sk-item {position: relative;z-index: 1;}#sk-container-id-4 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-4 div.sk-item::before, #sk-container-id-4 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-4 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-4 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-4 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-4 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-4 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-4 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-4 div.sk-label-container {text-align: center;}#sk-container-id-4 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-4 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-4\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>ExplainableBoostingClassifier(inner_bags=25, interactions=0, max_rounds=20000,\n",
       "                              n_jobs=-1, outer_bags=100)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-4\" type=\"checkbox\" checked><label for=\"sk-estimator-id-4\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ExplainableBoostingClassifier</label><div class=\"sk-toggleable__content\"><pre>ExplainableBoostingClassifier(inner_bags=25, interactions=0, max_rounds=20000,\n",
       "                              n_jobs=-1, outer_bags=100)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "ExplainableBoostingClassifier(inner_bags=25, interactions=0, max_rounds=20000,\n",
       "                              n_jobs=-1, outer_bags=100)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "EBM_S = ExplainableBoostingClassifier(interactions=0,max_rounds=20000,early_stopping_rounds=50,inner_bags=25,outer_bags=100,n_jobs=-1)#max_bins=256#,inner_bags=100,outer_bags=100 # 10\n",
    "EBM_S.fit(X_train_, Y_train_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "8d47b97d-c5e7-4289-844d-3835bac01bed",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--AUCROC perf = 0.7950137362637363 -- AUCPR perf = 0.7775959851942762\n"
     ]
    }
   ],
   "source": [
    "pred_ebm_s = EBM_S.predict_proba(X_test_)[:,1]\n",
    "best_aucroc_ebm_s = roc_auc_score(Y_test_, pred_ebm_s) \n",
    "best_aurpr_ebm_s = average_precision_score(Y_test_, pred_ebm_s) \n",
    "print(f\"--AUCROC perf = {best_aucroc_ebm_s} -- AUCPR perf = {best_aurpr_ebm_s}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "2b4e4473-4f21-4bc0-8eb1-b96fa4ef3287",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 869 ms, sys: 35.6 ms, total: 905 ms\n",
      "Wall time: 43.3 s\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-5 {color: black;background-color: white;}#sk-container-id-5 pre{padding: 0;}#sk-container-id-5 div.sk-toggleable {background-color: white;}#sk-container-id-5 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-5 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-5 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-5 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-5 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-5 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-5 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-5 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-5 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-5 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-5 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-5 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-5 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-5 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-5 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-5 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-5 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-5 div.sk-item {position: relative;z-index: 1;}#sk-container-id-5 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-5 div.sk-item::before, #sk-container-id-5 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-5 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-5 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-5 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-5 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-5 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-5 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-5 div.sk-label-container {text-align: center;}#sk-container-id-5 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-5 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-5\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>ExplainableBoostingClassifier(inner_bags=25, max_rounds=20000, n_jobs=-1,\n",
       "                              outer_bags=100)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-5\" type=\"checkbox\" checked><label for=\"sk-estimator-id-5\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ExplainableBoostingClassifier</label><div class=\"sk-toggleable__content\"><pre>ExplainableBoostingClassifier(inner_bags=25, max_rounds=20000, n_jobs=-1,\n",
       "                              outer_bags=100)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "ExplainableBoostingClassifier(inner_bags=25, max_rounds=20000, n_jobs=-1,\n",
       "                              outer_bags=100)"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "EBM = ExplainableBoostingClassifier(interactions=10,max_rounds=20000,early_stopping_rounds=50,inner_bags=25,outer_bags=100,n_jobs=-1)#max_bins=256#,inner_bags=100,outer_bags=100 # 10\n",
    "EBM.fit(X_train_, Y_train_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "bca16629-62e3-4d95-8ae3-e13b01448faa",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--AUCROC perf = 0.7982097069597069 -- AUCPR perf = 0.7826232899324753\n"
     ]
    }
   ],
   "source": [
    "pred_ebm = EBM.predict_proba(X_test_)[:,1]\n",
    "best_aucroc_ebm = roc_auc_score(Y_test_, pred_ebm) \n",
    "best_aurpr_ebm = average_precision_score(Y_test_, pred_ebm) \n",
    "print(f\"--AUCROC perf = {best_aucroc_ebm} -- AUCPR perf = {best_aurpr_ebm}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65e6d92f-cc9a-4044-9285-2d68fced29eb",
   "metadata": {},
   "source": [
    "## Explanations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "8403868b-365a-45be-974d-559edbeea879",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 67.5 ms, sys: 253 µs, total: 67.7 ms\n",
      "Wall time: 5.8 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "attributions_linear = pd.DataFrame(LR.get_feature_attribution(X_test_.values))\n",
    "attributions_linear.columns = feature_names\n",
    "attributions_linear['pred_proba'] = pred_lr\n",
    "attributions_linear['label'] = Y_test_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "284e53e7-8685-4943-8fef-a88e05c80645",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 45.6 ms, sys: 922 µs, total: 46.5 ms\n",
      "Wall time: 5.06 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "attributions_tabsra = pd.DataFrame(TabSRA.get_feature_attribution(X_test_.values))\n",
    "attributions_tabsra.columns = feature_names\n",
    "attributions_tabsra['pred_proba'] = pred_tabsra\n",
    "attributions_tabsra['label'] = Y_test_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "be45603f-aa52-4c55-ba40-66f76ded51e0",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "ntree_limit is deprecated, use `iteration_range` or model slicing instead.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 5.13 s, sys: 4.15 ms, total: 5.13 s\n",
      "Wall time: 626 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "explainer = shap.TreeExplainer(model_xgb)\n",
    "shap_values = explainer(X_test_)\n",
    "attributions_xgb = pd.DataFrame(shap_values.values)\n",
    "attributions_xgb.columns = feature_names\n",
    "attributions_xgb['pred_proba'] = pred_xgb\n",
    "attributions_xgb['label'] = Y_test_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "b0649f71-82f6-4b81-815f-3e42762c879a",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_feature_attribution_EBM(model, X_test):\n",
    "    tmp  = model.explain_local(X_test)\n",
    "    df_EBM = []\n",
    "    for i in range(len(X_test)):\n",
    "        df_EBM.append(tmp.data(i)['scores'])\n",
    "    df_EBM = pd.DataFrame(df_EBM)\n",
    "    df_EBM.columns = tmp.data(0)['names']\n",
    "    return df_EBM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "fa91b4e2-2ab6-4666-9a89-a625aaad9079",
   "metadata": {},
   "outputs": [],
   "source": [
    "attributions_EBM_S = get_feature_attribution_EBM(EBM_S, X_test_)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "d76556ac-c93d-4f2e-8b52-6e1f7f0806e7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 74.9 ms, sys: 7.91 ms, total: 82.8 ms\n",
      "Wall time: 81.1 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "attributions_EBM_S = get_feature_attribution_EBM(EBM_S, X_test_)\n",
    "feature_names_EBM_S = list(attributions_EBM_S.columns)\n",
    "attributions_EBM_S['pred_proba'] = pred_ebm_s\n",
    "attributions_EBM_S['label'] = Y_test_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "984a56d2-a57d-468e-b87f-24972c2a0c7f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 84.9 ms, sys: 8.04 ms, total: 93 ms\n",
      "Wall time: 91.1 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "attributions_EBM = get_feature_attribution_EBM(EBM, X_test_)\n",
    "feature_names_EBM = list(attributions_EBM.columns)\n",
    "attributions_EBM['pred_proba'] = pred_ebm\n",
    "attributions_EBM['label'] = Y_test_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "cbab4691-d8e7-46bd-81a5-9910cbb056b2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1076"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# sample random observations\n",
    "np.random.seed(42)\n",
    "random_ = np.random.randn(len(X_test_))\n",
    "sample_condition = np.where(random_>0, True, False)\n",
    "sample_condition.sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "351cdc83-d870-4497-920e-5386ace257b0",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test_tmp = X_test_[sample_condition]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "8ccad87c-2543-471f-a07f-494dbf765c0e",
   "metadata": {},
   "outputs": [],
   "source": [
    "alpha_stp = 1e-2\n",
    "n_pertub = 100\n",
    "epsilon_min = 1e-3\n",
    "n_features = len(feature_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "74d29dda-946c-4961-a0d6-2fa6c090cd74",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(42)\n",
    "tmp_x =np.random.normal(0,1*alpha_stp, n_features*n_pertub).reshape(n_pertub,n_features)\n",
    "rapport_denom  = np.linalg.norm(tmp_x,ord=2, axis=1)\n",
    "n_point = len(X_test_tmp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "df4811dd-bd2b-45ee-9764-71575a58afb7",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test_tmp_np   = X_test_tmp[feature_names].values\n",
    "X_explain_lr,y_explain_lr = attributions_linear[sample_condition][feature_names].values, attributions_linear[sample_condition].pred_proba.values\n",
    "X_explain_tabsra,y_explain_tabsra = attributions_tabsra[sample_condition][feature_names].values, attributions_tabsra[sample_condition].pred_proba.values\n",
    "X_explain_xgb,y_explain_xgb = attributions_xgb[sample_condition][feature_names].values, attributions_xgb[sample_condition].pred_proba.values\n",
    "X_explain_EBM_S,y_explain_EBM_S = attributions_EBM_S[sample_condition][feature_names_EBM_S].values, attributions_EBM_S[sample_condition].pred_proba.values\n",
    "X_explain_EBM,y_explain_EBM= attributions_EBM[sample_condition][feature_names_EBM].values, attributions_EBM[sample_condition].pred_proba.values\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "f78bbe1b-05e4-42bc-9109-b9e4af2e5a3f",
   "metadata": {},
   "outputs": [],
   "source": [
    "time_lr =[]\n",
    "for i in range(10):\n",
    "    start_ = time.time()\n",
    "    relative_norm_lr = np.zeros(n_point)\n",
    "    relative_pred_lr = np.zeros(n_point)\n",
    "    for i in range(n_point):\n",
    "        #tmp_x =np.random.normal(0,1*alpha_stp, 5*100).reshape(100,5)\n",
    "        X_new_np = X_test_tmp_np[i]+tmp_x\n",
    "        X_new = torch.Tensor(X_new_np)\n",
    "\n",
    "        y_pred_linear = LR.predict_proba(X_new)[:,1]\n",
    "        X_explain_new_ = np.linalg.norm(X_explain_lr[i] - LR.get_feature_attribution(X_new),ord=2,axis=1)\n",
    "\n",
    "        tmp = np.abs(y_explain_lr[i]-y_pred_linear)\n",
    "        ind_adversial = np.argmax(tmp)\n",
    "        relative_pred_lr[i] = tmp[ind_adversial]\n",
    "        relative_norm_lr[i] =  (X_explain_new_/rapport_denom)[ind_adversial]\n",
    "    time_lr.append(time.time()-start_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "8cff9aca-7cd8-4479-b1eb-52f4161d4d87",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.867487096786499, 0.024375846386603978)"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(time_lr),np.std(time_lr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fc35edf4-4176-4c45-aa1f-0d5cec12833a",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "10b46366-efe6-4771-8378-25181615bf1f",
   "metadata": {},
   "outputs": [],
   "source": [
    "time_tabsra =[]\n",
    "for i in range(10):\n",
    "    start_ = time.time()\n",
    "    relative_norm_tabsra = np.zeros(n_point)\n",
    "    relative_pred_tabsra = np.zeros(n_point)\n",
    "    for i in range(n_point):\n",
    "        #tmp_x =np.random.normal(0,1*alpha_stp, 5*100).reshape(100,5)\n",
    "        X_new_np = X_test_tmp_np[i]+tmp_x\n",
    "        X_new = torch.Tensor(X_new_np)\n",
    "\n",
    "        y_pred_tabsra = TabSRA.predict_proba(X_new)[:,1]\n",
    "        X_explain_new_ = np.linalg.norm(X_explain_tabsra[i] - TabSRA.get_feature_attribution(X_new),ord=2,axis=1)\n",
    "\n",
    "        tmp = np.abs(y_explain_tabsra[i]-y_pred_tabsra)\n",
    "        ind_adversial = np.argmax(tmp)\n",
    "        relative_pred_tabsra[i] = tmp[ind_adversial]\n",
    "        relative_norm_tabsra[i] =  (X_explain_new_/rapport_denom)[ind_adversial]\n",
    "    time_tabsra.append(time.time()-start_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "d7b29204-2b08-403b-b289-7f6000557fca",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1.459298062324524, 0.019078239796916197)"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(time_tabsra),np.std(time_tabsra)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "bb676724-89e4-4ff1-b404-71e193eb0779",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "ntree_limit is deprecated, use `iteration_range` or model slicing instead.\n"
     ]
    }
   ],
   "source": [
    "time_xgb =[]\n",
    "for i in range(10):\n",
    "    start_ = time.time()\n",
    "    relative_norm_xgb = np.zeros(n_point)\n",
    "    relative_pred_xgb = np.zeros(n_point)\n",
    "    for i in range(n_point):\n",
    "        #tmp_x =np.random.normal(0,1*alpha_stp, 5*100).reshape(100,5)\n",
    "        X_new_np = X_test_tmp_np[i]+tmp_x\n",
    "\n",
    "\n",
    "        y_pred_xgb = model_xgb.predict_proba(X_new_np)[:,1]\n",
    "        X_explain_new_ = np.linalg.norm(X_explain_xgb[i] - explainer(X_new_np).values,ord=2,axis=1)\n",
    "\n",
    "        tmp = np.abs(y_explain_xgb[i]-y_pred_xgb)\n",
    "        ind_adversial = np.argmax(tmp)\n",
    "        relative_pred_xgb[i] = tmp[ind_adversial]\n",
    "        relative_norm_xgb[i] =  (X_explain_new_/rapport_denom)[ind_adversial]\n",
    "    time_xgb.append(time.time()-start_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "7cd704b2-63b6-4a2d-ac46-ab34fe3c6836",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(21.29535698890686, 0.939028836328257)"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(time_xgb),np.std(time_xgb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "ea4335c4-d6a2-4c55-a413-3277f2711e40",
   "metadata": {},
   "outputs": [],
   "source": [
    "time_EBM_S =[]\n",
    "for i in range(10):\n",
    "    start_ = time.time()\n",
    "    relative_norm_EBM_S = np.zeros(n_point)\n",
    "    relative_pred_EBM_S = np.zeros(n_point)\n",
    "    for i in range(n_point):\n",
    "        #tmp_x =np.random.normal(0,1*alpha_stp, 5*100).reshape(100,5)\n",
    "        X_new_np = X_test_tmp_np[i]+tmp_x\n",
    "        X_new_df = pd.DataFrame(X_new_np,columns=feature_names)\n",
    "\n",
    "        y_pred_EBM_S = EBM_S.predict_proba(X_new_np)[:,1]\n",
    "        X_explain_EBM_S_ = np.linalg.norm(X_explain_EBM_S[i] - get_feature_attribution_EBM(EBM_S, X_new_df).values,ord=2,axis=1)\n",
    "\n",
    "        tmp = np.abs(y_explain_EBM_S[i]-y_pred_EBM_S)\n",
    "        ind_adversial = np.argmax(tmp)\n",
    "        relative_pred_EBM_S[i] = tmp[ind_adversial]\n",
    "        relative_norm_EBM_S[i] =  (X_explain_new_/rapport_denom)[ind_adversial]\n",
    "time_EBM_S.append(time.time()-start_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "58b17c74-9567-4203-b7b6-c68fced6cf9e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5.6853766441345215, 0.0)"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(time_EBM_S),np.std(time_EBM_S)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "07204d86-1264-4462-bf09-3cb25c6838fc",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "aa84ff0e-67b1-4ae4-9e54-c8365ea2abd9",
   "metadata": {},
   "outputs": [],
   "source": [
    "time_EBM =[]\n",
    "for i in range(10):\n",
    "    start_ = time.time()\n",
    "    relative_norm_EBM = np.zeros(n_point)\n",
    "    relative_pred_EBM = np.zeros(n_point)\n",
    "    for i in range(n_point):\n",
    "        #tmp_x =np.random.normal(0,1*alpha_stp, 5*100).reshape(100,5)\n",
    "        X_new_np = X_test_tmp_np[i]+tmp_x\n",
    "        X_new_df = pd.DataFrame(X_new_np,columns=feature_names)\n",
    "\n",
    "        y_pred_EBM = EBM.predict_proba(X_new_np)[:,1]\n",
    "        X_explain_EBM_ = np.linalg.norm(X_explain_EBM[i] - get_feature_attribution_EBM(EBM, X_new_df).values,ord=2,axis=1)\n",
    "\n",
    "        tmp = np.abs(y_explain_EBM[i]-y_pred_EBM)\n",
    "        ind_adversial = np.argmax(tmp)\n",
    "        relative_pred_EBM[i] = tmp[ind_adversial]\n",
    "        relative_norm_EBM[i] =  (X_explain_new_/rapport_denom)[ind_adversial]\n",
    "    time_EBM.append(time.time()-start_)       "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "c1ad0895-fafc-4482-8563-4efa84452987",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(6.661557221412659, 0.11419476946040463)"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(time_EBM),np.std(time_EBM)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "8c771d43-c298-4b55-98ce-ea37403950c6",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_classic_expl = pd.DataFrame()\n",
    "\n",
    "df_classic_expl[\"LR\"] = relative_norm_lr\n",
    "df_classic_expl[\"TabSRALinear\"] = relative_norm_tabsra\n",
    "df_classic_expl[\"EBM\"] = relative_norm_EBM\n",
    "df_classic_expl[\"EBM_S\"] = relative_norm_EBM_S\n",
    "df_classic_expl[\"XGBoost+TreeSHAP\"] = relative_norm_xgb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "ad79c576-b13a-4e0f-bc3a-d97ab381a71d",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_classic_output = pd.DataFrame()\n",
    "df_classic_output[\"LR\"] = relative_pred_lr\n",
    "df_classic_output[\"TabSRALinear\"] = relative_pred_tabsra\n",
    "df_classic_output[\"EBM\"] = relative_pred_EBM\n",
    "df_classic_output[\"EBM_S\"] = relative_pred_EBM_S\n",
    "df_classic_output[\"XGBoost+TreeSHAP\"] = relative_pred_xgb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "id": "c4d22c0f-77ff-4b5b-bde4-55474569d118",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>LR</th>\n",
       "      <th>TabSRALinear</th>\n",
       "      <th>EBM</th>\n",
       "      <th>EBM_S</th>\n",
       "      <th>XGBoost+TreeSHAP</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.307179</td>\n",
       "      <td>0.467895</td>\n",
       "      <td>1.147138</td>\n",
       "      <td>1.147138</td>\n",
       "      <td>0.454984</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.307179</td>\n",
       "      <td>0.399638</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.990627</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.307179</td>\n",
       "      <td>0.179733</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.517604</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.307179</td>\n",
       "      <td>0.317978</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.235470</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.307179</td>\n",
       "      <td>0.298078</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.082124</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         LR  TabSRALinear       EBM     EBM_S  XGBoost+TreeSHAP\n",
       "0  0.307179      0.467895  1.147138  1.147138          0.454984\n",
       "1  0.307179      0.399638  0.000000  0.000000          0.990627\n",
       "2  0.307179      0.179733  0.000000  0.000000          0.517604\n",
       "3  0.307179      0.317978  0.000000  0.000000          0.235470\n",
       "4  0.307179      0.298078  0.000000  0.000000          2.082124"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_classic_expl.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "1398f4e2-b0d9-4c85-95b4-852e1df64a2b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1076, 2092)"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_classic_expl),len(X_test_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "f2f22d2b-785b-4f9b-9e66-5f5466dba3a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_classic_expl.columns = ['LR', 'SRA', 'EBM', 'EBM_S', 'XGB_SHAP']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "acd3fc95-227b-4dde-b0b0-519047de5f14",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(data=df_classic_expl)\n",
    "plt.ylabel(\"Lipschitz Estimate\",weight = 'bold',fontsize=14)\n",
    "plt.xticks(weight = 'bold',fontsize=11,rotation=0)\n",
    "plt.yticks(weight = 'bold',fontsize=14)\n",
    "plt.savefig('Stability_HelocFico_hideen1_1E_2.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "id": "91166030-a9d1-470e-9d60-8eed6ce7e775",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_classic_output.columns = ['LR', 'SRA', 'EBM', 'EBM_S', 'XGB_SHAP']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "id": "90436108-b37c-4b4a-9c0a-0adb3e764c9c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(data=df_classic_output)\n",
    "#plt.ylabel(\"Output Change\",weight = 'bold',fontsize=14)\n",
    "plt.xticks(weight = 'bold',fontsize=11,rotation=0)\n",
    "plt.yticks(weight = 'bold',fontsize=14)\n",
    "plt.savefig('Stability_OutputChange_HelocFico_hideen1_1E_2.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d5808f1e-87c8-4c46-b049-c03a2ce748a4",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}