02_base_model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "1bad8144-136d-40d9-aed8-287e647aafce",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Modify the file 00_setup.py to define input/output file paths on your system\n",
    "# The information in 00_setup.py will be used across notebooks\n",
    "from importlib.machinery import SourceFileLoader\n",
    "setup = SourceFileLoader(\"setup\", \"./00_setup.py\").load_module()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dc026528-466a-4f4d-8264-988914dec40c",
   "metadata": {},
   "source": [
    "# 02: Base Model\n",
    "Train a model.  Keep the model fairly small to make this script, and future ones which use computationally intensive methods like SHAP, feasible to run on home hardware."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "45bb1b82-0d0e-420a-afbc-758aeed6a5fc",
   "metadata": {},
   "source": [
    "Due to compatibility with sklearn_gbmi, I am using a GradientBoostingClassifier model.  This model is slow-ish and does not seem as accurate as, say, xgboost.  However, since I am interested in interactions, this is OK."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "14c09328-957e-46c8-95cb-74a5fef75ddf",
   "metadata": {},
   "source": [
    "*This script takes about 15 minutes to run on my MacBook Air*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "3f2c5433-431e-4c87-87f6-60c4323ace58",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pickle\n",
    "from pathlib import Path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "7c6b208c-7072-4076-bca2-579b7c71a762",
   "metadata": {},
   "outputs": [],
   "source": [
    "import sklearn as sk\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from sklearn.metrics import accuracy_score, f1_score, precision_score, recall_score, \\\n",
    "    average_precision_score, roc_auc_score\n",
    "from sklearn.inspection import permutation_importance\n",
    "from scikitplot.metrics import plot_lift_curve\n",
    "from sklearn.ensemble import GradientBoostingClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "10e13dd1-8716-43e5-9f2a-03ef0848cce0",
   "metadata": {},
   "outputs": [],
   "source": [
    "from interactions_package.interactions_package import ip_plot\n",
    "ip_plot.plot_defaults()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "88ac8af9-6049-49cf-84e8-1929fb5b4635",
   "metadata": {},
   "source": [
    "## Import Processed Datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "f2b638b7-5a92-4568-ab16-f1d294972bc0",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df = pd.read_parquet(Path(setup.temp_path).joinpath('01_DATA_combined_base_train.parquet'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "56da9346-4d6a-480f-8406-0114f59bdeca",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_df = pd.read_parquet(Path(setup.temp_path).joinpath('01_DATA_combined_base_test.parquet'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "0afcb56a-855b-417e-b5f7-8da2ad7d8649",
   "metadata": {},
   "outputs": [],
   "source": [
    "val_df = pd.read_parquet(Path(setup.temp_path).joinpath('01_DATA_combined_base_val.parquet'))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5845461a-7661-46d6-9f9c-0f1242b560e9",
   "metadata": {},
   "source": [
    "##### Get predictors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "a3077f2a-7a81-44dd-8329-cc0aaa989ff2",
   "metadata": {},
   "outputs": [],
   "source": [
    "predictor_features = pd.read_csv(Path(setup.temp_path) \\\n",
    "                                 .joinpath('01_REPORT_predictor_features.csv'))['feature'].tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "9581e0ad-f8c6-405b-b0cb-e6953566cd9b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "18"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(set(predictor_features))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "14c43fa1-0370-419a-a58a-7b097c533239",
   "metadata": {},
   "source": [
    "## Model Hyperparameter Tune and Fit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "ed50581a-29bd-4f1f-8f31-428792373c16",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.19962582237312326"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_target = np.mean(train_df['target'])\n",
    "mean_target"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c3d6c02b-0612-4b77-8c7b-bf49db846026",
   "metadata": {},
   "source": [
    "##### Do a simple hyperparameter search"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "143db5cd-7086-4449-859a-0dda5f0de4c0",
   "metadata": {},
   "outputs": [],
   "source": [
    "base_model = GradientBoostingClassifier(n_estimators = 20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "860d5178-5f8c-4b43-b428-8b65bd72f21c",
   "metadata": {},
   "outputs": [],
   "source": [
    "hyperparam_serarch = {'learning_rate': [0.01, 0.05, 0.1],\n",
    "                      'min_samples_split': [1, 10, 20],\n",
    "                      'min_samples_leaf': [1, 5, 20],\n",
    "                      'max_depth': [4, 5, 6],\n",
    "                      'max_leaf_nodes': [None, 30, 100]}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "8ad6cdac-6748-4833-9cdf-1aca3b127b31",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set up a cross validation without the final fit; don't use early stopping yet\n",
    "random_search = RandomizedSearchCV(base_model, \n",
    "                                   hyperparam_serarch,\n",
    "                                   n_iter=20, \n",
    "                                   n_jobs=-1,\n",
    "                                   random_state=1055,\n",
    "                                   verbose = 0,\n",
    "                                   refit = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "0ba1c884-3b77-4571-babc-de5512d295ba",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Do the cross validation parameter search\n",
    "rs_fit = random_search.fit(train_df[predictor_features], train_df['target']);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "2f23315e-daf5-4a9f-b649-959fa9610f55",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'min_samples_split': 20, 'min_samples_leaf': 20, 'max_leaf_nodes': None, 'max_depth': 6, 'learning_rate': 0.1}\n"
     ]
    }
   ],
   "source": [
    "print(random_search.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "11a328c1-b494-4050-b4a4-601b067ff35b",
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.DataFrame({k:[random_search.best_params_[k]] for k in random_search.best_params_.keys()}) \\\n",
    "    .to_csv(Path(setup.temp_path).joinpath('02_REPORT_selected_features.csv'), index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "23632fbe-e27c-494a-ac5c-1fa318d78f11",
   "metadata": {},
   "source": [
    "##### Do a final fit with the full training data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "90e99663-1ed5-4e21-a994-5e1d3bc59a6d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 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-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 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-1 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-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 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-1 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-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 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-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 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-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 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-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GradientBoostingClassifier(max_depth=6, min_samples_leaf=20,\n",
       "                           min_samples_split=20, n_estimators=20)</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-1\" type=\"checkbox\" checked><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GradientBoostingClassifier</label><div class=\"sk-toggleable__content\"><pre>GradientBoostingClassifier(max_depth=6, min_samples_leaf=20,\n",
       "                           min_samples_split=20, n_estimators=20)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "GradientBoostingClassifier(max_depth=6, min_samples_leaf=20,\n",
       "                           min_samples_split=20, n_estimators=20)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "base_model.set_params(**random_search.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "046a6a48-0035-43cb-aff6-3e983bbb3bef",
   "metadata": {},
   "outputs": [
    {
     "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>GradientBoostingClassifier(max_depth=6, min_samples_leaf=20,\n",
       "                           min_samples_split=20, n_estimators=20)</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\">GradientBoostingClassifier</label><div class=\"sk-toggleable__content\"><pre>GradientBoostingClassifier(max_depth=6, min_samples_leaf=20,\n",
       "                           min_samples_split=20, n_estimators=20)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "GradientBoostingClassifier(max_depth=6, min_samples_leaf=20,\n",
       "                           min_samples_split=20, n_estimators=20)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "base_model.fit(train_df[predictor_features], train_df['target'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "101530fd-61a1-4786-bdcb-7a0103f2fa78",
   "metadata": {},
   "source": [
    "##### Save the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "407441c5-cb57-4f3e-aab3-01f0532112cf",
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(Path(setup.temp_path).joinpath('02_MODEL_base_gc.pkl'), 'wb') as po:\n",
    "    pickle.dump(base_model, po)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "0a2274fd-bf58-4f0c-8694-caef65c5bf6e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'ccp_alpha': 0.0,\n",
       " 'criterion': 'friedman_mse',\n",
       " 'init': None,\n",
       " 'learning_rate': 0.1,\n",
       " 'loss': 'log_loss',\n",
       " 'max_depth': 6,\n",
       " 'max_features': None,\n",
       " 'max_leaf_nodes': None,\n",
       " 'min_impurity_decrease': 0.0,\n",
       " 'min_samples_leaf': 20,\n",
       " 'min_samples_split': 20,\n",
       " 'min_weight_fraction_leaf': 0.0,\n",
       " 'n_estimators': 20,\n",
       " 'n_iter_no_change': None,\n",
       " 'random_state': None,\n",
       " 'subsample': 1.0,\n",
       " 'tol': 0.0001,\n",
       " 'validation_fraction': 0.1,\n",
       " 'verbose': 0,\n",
       " 'warm_start': False}"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "base_model.get_params()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3e6c7c91-a7a2-4f09-89f7-825b1d35cf8f",
   "metadata": {},
   "source": [
    "## Metrics"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d6e46f35-6ad5-43fe-bb1d-934fdfcab3ca",
   "metadata": {},
   "source": [
    "##### Classification metrics\n",
    "Accuracy, f1, precision, recall"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "59148076-e3e9-41a3-a900-db1851e91e5f",
   "metadata": {},
   "outputs": [],
   "source": [
    "metrics_class_df = pd.concat([pd.DataFrame({metf.__name__: [metf(df['target'], \n",
    "                                                                 base_model.predict(df[predictor_features]))] \\\n",
    "              for metf in [accuracy_score, f1_score, precision_score, recall_score]})\n",
    "           for df in [train_df, test_df, val_df]], keys = ['train', 'test', 'val']) \\\n",
    "    .reset_index(level=1, drop=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e3d4d175-a65a-490f-9b74-fe9357b95688",
   "metadata": {},
   "source": [
    "##### AUC scores\n",
    "PR-AUC and ROC-AUC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "08a44162-1f9d-48c6-a7e5-e349f3df8bf6",
   "metadata": {},
   "outputs": [],
   "source": [
    "metrics_auc_df = pd.concat([pd.DataFrame({metf.__name__: [metf(df['target'], \n",
    "                                                               base_model.predict_proba(df[predictor_features])[:,1])] \\\n",
    "              for metf in [average_precision_score, roc_auc_score]})\n",
    "           for df in [train_df, test_df, val_df]], keys = ['train', 'test', 'val']) \\\n",
    "    .reset_index(level=1, drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "fb4f2bae-37b3-464d-915b-1ab7ca15ae86",
   "metadata": {},
   "outputs": [],
   "source": [
    "metrics_df = pd.concat([metrics_class_df, metrics_auc_df], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "4b889110-286b-4db5-8083-94de96767204",
   "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>accuracy_score</th>\n",
       "      <th>f1_score</th>\n",
       "      <th>precision_score</th>\n",
       "      <th>recall_score</th>\n",
       "      <th>average_precision_score</th>\n",
       "      <th>roc_auc_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>train</th>\n",
       "      <td>0.801462</td>\n",
       "      <td>0.028102</td>\n",
       "      <td>0.616859</td>\n",
       "      <td>0.014379</td>\n",
       "      <td>0.377591</td>\n",
       "      <td>0.712000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>test</th>\n",
       "      <td>0.801172</td>\n",
       "      <td>0.026136</td>\n",
       "      <td>0.588063</td>\n",
       "      <td>0.013365</td>\n",
       "      <td>0.372586</td>\n",
       "      <td>0.709407</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>val</th>\n",
       "      <td>0.801406</td>\n",
       "      <td>0.026850</td>\n",
       "      <td>0.615752</td>\n",
       "      <td>0.013724</td>\n",
       "      <td>0.364952</td>\n",
       "      <td>0.702959</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       accuracy_score  f1_score  precision_score  recall_score  \\\n",
       "train        0.801462  0.028102         0.616859      0.014379   \n",
       "test         0.801172  0.026136         0.588063      0.013365   \n",
       "val          0.801406  0.026850         0.615752      0.013724   \n",
       "\n",
       "       average_precision_score  roc_auc_score  \n",
       "train                 0.377591       0.712000  \n",
       "test                  0.372586       0.709407  \n",
       "val                   0.364952       0.702959  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metrics_df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "438d53af-afc1-4739-b16d-5f13ad69a992",
   "metadata": {},
   "source": [
    "This model isn't terribly good.  The target rate is ~20% and so 0.20 is the baseline PR-AUC (average precision score).  So the model is about 1.8 x random.  However I am interested in interactions and want to keep this model simple so this is OK for this purpose\n",
    "\n",
    "Looking at precision and recall, the max f1 threshold might not be the best decision threshold.  However I am not using a decision threshold and so can ignore this."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cd9623a3-a9e3-4147-9a02-a0ae4a63d0b0",
   "metadata": {},
   "source": [
    "## Lift"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "090ffb3d-ea76-4d69-89af-8d1675ef5971",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 400x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_lift_curve(val_df['target'], \n",
    "                base_model.predict_proba(val_df[predictor_features]), \n",
    "                title='Lift Curve', ax=None, figsize=(4,3), \n",
    "                title_fontsize='large', text_fontsize='medium')\n",
    "plt.gca().legend(loc='upper right')\n",
    "plt.gcf().savefig(Path(setup.temp_path).joinpath('02_PLOT_lift_curve.png'),\n",
    "                 bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b015e723-c81c-441b-9d8d-024fc4987e30",
   "metadata": {},
   "source": [
    "## Feature Importance "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3074e8f2-ef62-4592-b546-391bcc5182c9",
   "metadata": {},
   "source": [
    "##### Impurity Importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "fa90f824-4957-4a9e-8e91-f76610e2fd3c",
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_importance_df= pd.DataFrame({'feature': predictor_features,\n",
    "                                    'importance': base_model.feature_importances_}) \\\n",
    "    .sort_values('importance', ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "2d4d8f93-16bc-4a82-bc80-facda1390996",
   "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>feature</th>\n",
       "      <th>importance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>int_rate</td>\n",
       "      <td>0.716283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>term</td>\n",
       "      <td>0.080824</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>mort_acc</td>\n",
       "      <td>0.033547</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     feature  importance\n",
       "2   int_rate    0.716283\n",
       "1       term    0.080824\n",
       "11  mort_acc    0.033547"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature_importance_df.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "480a593c-3a30-4118-abbc-9a7b8b408e55",
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_importance_df.to_csv(Path(setup.temp_path).joinpath('02_REPORT_importance_impurity.csv'), index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "ccb3fb93-ddee-47f7-b9da-5aba2081dd52",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "feature_importance_df.head(10).set_index('feature') \\\n",
    "    .plot(kind='barh', legend=None, color='lightslategray')\n",
    "plt.gca().invert_yaxis()\n",
    "plt.xlabel('impurity importance')\n",
    "plt.ylabel(None)\n",
    "plt.gcf().savefig(Path(setup.temp_path).joinpath('02_PLOT_importance_gain.png'),\n",
    "                 bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "18aa7d09-2cfe-474f-b185-886aa3bd1b9e",
   "metadata": {},
   "source": [
    "##### Permutation Importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "ceb6b6ea-513d-4a59-b0ab-9bf6a91787be",
   "metadata": {},
   "outputs": [],
   "source": [
    "perm_importance = permutation_importance(base_model, \n",
    "                                         train_df[predictor_features], \n",
    "                                         train_df['target'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "17740501-02fb-4e62-a088-b92c73d7aebb",
   "metadata": {},
   "outputs": [],
   "source": [
    "perm_importance_df = pd.DataFrame({'features':predictor_features,\n",
    "                                   'importance': perm_importance['importances_mean']}) \n",
    "perm_importance_df['abs_importance'] = np.abs(perm_importance_df['importance'])\n",
    "perm_importance_df = perm_importance_df.sort_values('abs_importance', ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "6c0e0cf8-b1a8-4ed3-bbfe-4411e479d6cf",
   "metadata": {},
   "outputs": [],
   "source": [
    "perm_importance_df \\\n",
    "    .to_csv(Path(setup.temp_path).joinpath('02_DATA_importance_perturb.csv'),\n",
    "            index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "0e429e9d-c224-4e50-8d80-d3e453739a2b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "perm_importance_df.head(10) \\\n",
    "    .plot(kind='barh', x='features', y='importance', legend=None, color='lightslategray')\n",
    "plt.gca().invert_yaxis()\n",
    "plt.xlabel('permutation importance')\n",
    "plt.ylabel(None)\n",
    "plt.gcf().savefig(Path(setup.temp_path).joinpath('02_PLOT_importance_perturb.png'),\n",
    "                 bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4f9e6a72-bd44-41a0-b56f-495165e90d87",
   "metadata": {},
   "source": [
    "I am not an expert on lending, but it makes sense that the interest rate is the most important feature.  Interest rate should capture financial risk, as opposed to strict default risk. So the interest rate should incorporate information beyond the default rate, such as loan amount.   "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f6cd1ca8-1547-444b-b826-a14a9f6d6c13",
   "metadata": {},
   "source": [
    "Interactions between interest rates and other features might capture the difference between financial and default risks.  Or, interactions could reflect sub-populations for whom the interst rate is not well calibrated."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "579b2434-9816-427c-9a89-306389c5e184",
   "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": "3.10.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}