82_groupings_shapley.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "44de2119-12ef-435f-a83d-4ed2ae732eef",
   "metadata": {},
   "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": "43809f6b-4858-4b27-a7bb-835e2406b65d",
   "metadata": {},
   "source": [
    "# SHAP explorations of mean encoding features\n",
    "In hierarchical encoding, more general NAICS categories are worse than specific.  For DGI clusters, it's the opposite.  Why? "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1bfe56ac-16e2-4c30-b39f-9127cbeb6795",
   "metadata": {},
   "source": [
    "Here, I look at SHAP values for the holdout cases for models with hiearchical encodings from DGI vs standard NAICS.  The question is whether the encoded feature is more \"important\" for one of the cases, and how the response varies with the value of the feature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "08819fb3-64c8-45bd-afab-e059cad57f9f",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from pathlib import Path\n",
    "import importlib, os, pickle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "2f93bd7e-941b-46e5-9494-bb878f4b0b28",
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "b2f37eae-8420-4677-8f7e-d35eeaa67b4c",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sba_gnn.sba_gnn import sg_plot \n",
    "sg_plot.plot_defaults()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "27c513ff-3803-4984-8d47-47b3ee19a905",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import shap\n",
    "import xgboost as xgb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "4c82779e-86f8-4d3a-afdc-d8a94d07abcc",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.ticker as mtick"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "487eaf88-0ec8-4bfe-8079-89ee297a1310",
   "metadata": {},
   "source": [
    "## Import Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "f9e6efbd-0f3f-4ac5-bb5a-7c977b5d418e",
   "metadata": {},
   "outputs": [],
   "source": [
    "sba_loans = pd.read_parquet(Path(setup.temp_path).joinpath('01_DATA_transformed.parquet'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "47cb8e6c-ada4-4836-a86b-f566d476d470",
   "metadata": {},
   "outputs": [],
   "source": [
    "embed_df = pd.read_parquet(Path(setup.temp_path).joinpath('63_DATA_embeddings_tsne_naics.parquet'))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "05a56c8e-b125-4dc2-ab75-767e3a8ec51d",
   "metadata": {
    "tags": []
   },
   "source": [
    "##### Hierarchical encodings only\n",
    "Map the DGI based encoded feature into the loans"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "08ddc76d-85ef-4ce9-8f0c-5a689f2c64eb",
   "metadata": {},
   "outputs": [],
   "source": [
    "hier_df = pd.read_parquet(Path(setup.temp_path).joinpath('65_DATA_dgi_mhier.parquet'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "24c3af10-2eaa-46b5-92ed-1b38d9ffea6e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['LoanNr_ChkDgt', 'dset', 'dset_naics_holdout', 'target', 'mhier_NAICS', 'menc_dgi', 'NAICS']\n"
     ]
    }
   ],
   "source": [
    "print(list(hier_df.columns))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "112f5543-adec-4c36-8ff3-5097bfd20ec3",
   "metadata": {},
   "outputs": [],
   "source": [
    "sba_loans = sba_loans.merge(hier_df[['LoanNr_ChkDgt','menc_dgi']],\n",
    "                            on='LoanNr_ChkDgt')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "9562b799-bff9-44a0-b5a7-e3ad8e9f4512",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(688081, 59)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sba_loans.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "aee601d9-2304-4b94-b83d-1ab03e8ecf44",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Encoding column into dictionary\n",
    "feature_dict = {'menc': 'menc_NAICS', 'naics':'mhier_NAICS',\n",
    "              'dgi': 'menc_dgi'}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "5e151a18-60aa-4108-b88c-9e9c43845d5b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Order for plots and tables\n",
    "feature_order = ['menc', 'naics', 'dgi']\n",
    "num_models = len(feature_order)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "7ffa2f65-18fd-41be-a97d-c1ed5b49b98c",
   "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>menc_NAICS</th>\n",
       "      <th>mhier_NAICS</th>\n",
       "      <th>menc_dgi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>menc_NAICS</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.962039</td>\n",
       "      <td>0.925959</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mhier_NAICS</th>\n",
       "      <td>0.962039</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.916195</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>menc_dgi</th>\n",
       "      <td>0.925959</td>\n",
       "      <td>0.916195</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             menc_NAICS  mhier_NAICS  menc_dgi\n",
       "menc_NAICS     1.000000     0.962039  0.925959\n",
       "mhier_NAICS    0.962039     1.000000  0.916195\n",
       "menc_dgi       0.925959     0.916195  1.000000"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sba_loans[feature_dict.values()].corr()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "79257bc3-18c7-43ed-8272-de062e3a425c",
   "metadata": {},
   "source": [
    "##### Models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "0662055b-002c-4808-8c4a-2a63c68e0563",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Standard target encoding\n",
    "xgb_model_menc = xgb.Booster()\n",
    "xgb_model_menc.load_model(Path(setup.temp_path).joinpath('04_MODEL_xgboost.json'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "2bc59e58-7be8-4fe2-aa64-ffc7a002f8d1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Hierarchical target encoding by NAICS\n",
    "xgb_model_naics = xgb.Booster()\n",
    "xgb_model_naics.load_model(Path(setup.temp_path).joinpath('10_MODEL_xgboost.json'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "48fde8e8-9d13-44a7-b97f-993f81cc071a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Hierarchical target encoding by DGI\n",
    "xgb_model_dgi = xgb.Booster()\n",
    "xgb_model_dgi.load_model(Path(setup.temp_path).joinpath('65_MODEL_xgboost.json'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "cb4f5086-f216-4ade-8178-3a0fea53c7f5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Models into a dictionary\n",
    "model_dict = {'menc': xgb_model_menc, 'naics':xgb_model_naics,\n",
    "              'dgi': xgb_model_dgi}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "18539d39-0b1c-4acb-b1b7-622da0696257",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Predictor dictionary for compactness\n",
    "predictor_dict = {k: setup.predictor_features + [feature_dict[k]]\n",
    "                 for k in feature_dict.keys()}"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e94cde77-1c98-45b1-ae46-42e98da38a4e",
   "metadata": {
    "tags": []
   },
   "source": [
    "## SHAP Explainers, Values\n",
    "For the 3 models, generate explainers then explanations"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a82f358a-8617-45f8-b186-d0b4b41130b9",
   "metadata": {
    "tags": []
   },
   "source": [
    "##### Explainers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "3f975d35-90b3-4f9a-9d8b-023e1f9f9687",
   "metadata": {},
   "outputs": [],
   "source": [
    "explainer_dict = {k: \\\n",
    "                  shap.TreeExplainer(model_dict[k],feature_names=predictor_dict[k]) \\\n",
    "              for k in feature_dict.keys()}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "f6fe2cf9-271d-4a54-b98e-3a858c026241",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'menc': <shap.explainers._tree.TreeExplainer at 0x2b3e78ac0>,\n",
       " 'naics': <shap.explainers._tree.TreeExplainer at 0x2b3e78c10>,\n",
       " 'dgi': <shap.explainers._tree.TreeExplainer at 0x2b3e78820>}"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "explainer_dict"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c4693875-7daa-431b-89c3-72e57ad3d7a9",
   "metadata": {},
   "source": [
    "##### Holdout values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "d167e0fb-5661-46d9-be1e-6ff1017558a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "holdout_df = sba_loans[sba_loans['dset_naics_holdout'] == 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "95420277-5ce4-484b-9710-e1d4a4012020",
   "metadata": {},
   "outputs": [],
   "source": [
    "shap_holdout_dict = {k: \n",
    "                     explainer_dict[k].shap_values(holdout_df[predictor_dict[k]])\n",
    "                     for k in feature_dict.keys()}"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "289b90c9-84cb-4dd8-92bf-9926c53556e7",
   "metadata": {},
   "source": [
    "##### Random test values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "5b9996b6-2f4e-480a-b504-e77721366006",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_nh_df = sba_loans[sba_loans['dset_naics_holdout'] == 0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "1b176f22-dee2-435a-8d2b-be6ab9c4c70a",
   "metadata": {},
   "outputs": [],
   "source": [
    "shap_test_dict = {k: \n",
    "                     explainer_dict[k].shap_values(test_nh_df[predictor_dict[k]])\n",
    "                     for k in feature_dict.keys()}"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c5c07132-7b5d-4cef-a7e0-e7cd1f9065b2",
   "metadata": {},
   "source": [
    "## SHAP Dataframes"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a78171a6-532b-4fda-a2e2-5999de85b661",
   "metadata": {},
   "source": [
    "##### Wide version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "7cee811a-abf1-46ed-b645-ec72a3ae53b5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Wide dataframe for a single model\n",
    "def get_shap_df(data, shap_values, predictor_features, enc_feature):\n",
    "    df = pd.DataFrame(shap_values, columns=predictor_features,index = data.index) \\\n",
    "        .rename(columns={enc_feature:'enc_shap'})\n",
    "    df = pd.concat([data[['LoanNr_ChkDgt', 'target', 'NAICS',\n",
    "                          enc_feature]].rename(columns={enc_feature:'enc_val'}),\n",
    "                    df], axis=1)\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "5f800356-19c6-4038-a083-9c71f06a6f16",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dataframe for multiple models\n",
    "def get_wide_df(data, shap_dict,\n",
    "               predictor_dict = predictor_dict,\n",
    "               feature_dict = feature_dict,\n",
    "               model_categories = feature_order):\n",
    "    df = pd.concat([get_shap_df(data, shap_dict[k], predictor_dict[k], feature_dict[k])\n",
    "                    for k in feature_dict.keys()],\n",
    "                   keys=feature_dict.keys()) \\\n",
    "        .reset_index(level=0) \\\n",
    "        .rename(columns={'level_0':'model'}) \\\n",
    "        .reset_index(drop=True)\n",
    "    \n",
    "    cat_type = pd.CategoricalDtype(categories=feature_order,ordered=True)\n",
    "    df['model'] =  df['model'].astype(cat_type)\n",
    "    \n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "db9e5591-dd88-4bdd-b23c-85aa9bde8a10",
   "metadata": {},
   "outputs": [],
   "source": [
    "holdout_wide_df = get_wide_df(holdout_df, shap_holdout_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "2570dcd8-84f1-4e25-948f-dbf503465739",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_wide_df = get_wide_df(test_nh_df, shap_test_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "f5a25963-f9ba-4c6b-bf67-fdba4d39947c",
   "metadata": {},
   "outputs": [],
   "source": [
    "wide_df = pd.concat([holdout_wide_df, test_wide_df],\n",
    "                    keys = [1, 0]) \\\n",
    "    .reset_index(level=0) \\\n",
    "    .rename(columns={'level_0':'dset_naics_holdout'}) \\\n",
    "    .reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "403065ab-18e2-41f9-881b-49847b23644b",
   "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>dset_naics_holdout</th>\n",
       "      <th>model</th>\n",
       "      <th>LoanNr_ChkDgt</th>\n",
       "      <th>target</th>\n",
       "      <th>NAICS</th>\n",
       "      <th>enc_val</th>\n",
       "      <th>NoEmp</th>\n",
       "      <th>CreateJob</th>\n",
       "      <th>LowDoc</th>\n",
       "      <th>DisbursementGross</th>\n",
       "      <th>new_business</th>\n",
       "      <th>urban_flag</th>\n",
       "      <th>franchise_flag</th>\n",
       "      <th>enc_shap</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>menc</td>\n",
       "      <td>1000146010</td>\n",
       "      <td>0</td>\n",
       "      <td>445299</td>\n",
       "      <td>0.204655</td>\n",
       "      <td>0.088108</td>\n",
       "      <td>-0.001880</td>\n",
       "      <td>0.017768</td>\n",
       "      <td>-0.149431</td>\n",
       "      <td>0.193137</td>\n",
       "      <td>0.170542</td>\n",
       "      <td>0.029723</td>\n",
       "      <td>0.091684</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>menc</td>\n",
       "      <td>1000734007</td>\n",
       "      <td>0</td>\n",
       "      <td>512110</td>\n",
       "      <td>0.204655</td>\n",
       "      <td>0.044254</td>\n",
       "      <td>-0.010897</td>\n",
       "      <td>-0.250198</td>\n",
       "      <td>0.049062</td>\n",
       "      <td>-0.053356</td>\n",
       "      <td>-0.995948</td>\n",
       "      <td>-0.003018</td>\n",
       "      <td>0.078733</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>menc</td>\n",
       "      <td>1000745006</td>\n",
       "      <td>0</td>\n",
       "      <td>541810</td>\n",
       "      <td>0.204655</td>\n",
       "      <td>-0.000311</td>\n",
       "      <td>-0.352729</td>\n",
       "      <td>0.042809</td>\n",
       "      <td>0.397000</td>\n",
       "      <td>-0.010701</td>\n",
       "      <td>-0.865773</td>\n",
       "      <td>0.000070</td>\n",
       "      <td>-0.021616</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   dset_naics_holdout model LoanNr_ChkDgt  target   NAICS   enc_val     NoEmp  \\\n",
       "0                   1  menc    1000146010       0  445299  0.204655  0.088108   \n",
       "1                   1  menc    1000734007       0  512110  0.204655  0.044254   \n",
       "2                   1  menc    1000745006       0  541810  0.204655 -0.000311   \n",
       "\n",
       "   CreateJob    LowDoc  DisbursementGross  new_business  urban_flag  \\\n",
       "0  -0.001880  0.017768          -0.149431      0.193137    0.170542   \n",
       "1  -0.010897 -0.250198           0.049062     -0.053356   -0.995948   \n",
       "2  -0.352729  0.042809           0.397000     -0.010701   -0.865773   \n",
       "\n",
       "   franchise_flag  enc_shap  \n",
       "0        0.029723  0.091684  \n",
       "1       -0.003018  0.078733  \n",
       "2        0.000070 -0.021616  "
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "wide_df.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "5e6863f2-7d7c-492c-b22c-e3d8cde4ad1d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "model\n",
       "menc     688081\n",
       "naics    688081\n",
       "dgi      688081\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "wide_df['model'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "65dd7a2e-5a3a-4192-91bc-9c6a98056792",
   "metadata": {},
   "outputs": [],
   "source": [
    "wide_df.to_parquet(Path(setup.temp_path).joinpath('82_DATA_shap_wide.parquet'))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d04c4d87-c16f-46ef-b352-fcd3cef9f821",
   "metadata": {},
   "source": [
    "##### Long version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "f37e7fb4-5627-4ec1-9c1d-a19390826138",
   "metadata": {},
   "outputs": [],
   "source": [
    "shap_cols = setup.predictor_features + ['enc_shap']\n",
    "id_cols = [c for c in wide_df.columns if c not in shap_cols]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "310707d3-ad83-4ca9-b8c4-0bf64893a250",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['dset_naics_holdout', 'model', 'LoanNr_ChkDgt', 'target', 'NAICS', 'enc_val']"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "id_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "37f8846e-3a86-470b-b9c4-d9fc0b3e4efe",
   "metadata": {},
   "outputs": [],
   "source": [
    "long_df = wide_df.melt(id_vars=id_cols)\n",
    "long_df['enc_feat'] = np.where(long_df['variable'] == 'enc_shap', 1, 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "c849c505-2bef-4105-a81f-12fc3fec43c3",
   "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>dset_naics_holdout</th>\n",
       "      <th>model</th>\n",
       "      <th>LoanNr_ChkDgt</th>\n",
       "      <th>target</th>\n",
       "      <th>NAICS</th>\n",
       "      <th>enc_val</th>\n",
       "      <th>variable</th>\n",
       "      <th>value</th>\n",
       "      <th>enc_feat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>menc</td>\n",
       "      <td>1000146010</td>\n",
       "      <td>0</td>\n",
       "      <td>445299</td>\n",
       "      <td>0.204655</td>\n",
       "      <td>NoEmp</td>\n",
       "      <td>0.088108</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>menc</td>\n",
       "      <td>1000734007</td>\n",
       "      <td>0</td>\n",
       "      <td>512110</td>\n",
       "      <td>0.204655</td>\n",
       "      <td>NoEmp</td>\n",
       "      <td>0.044254</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>menc</td>\n",
       "      <td>1000745006</td>\n",
       "      <td>0</td>\n",
       "      <td>541810</td>\n",
       "      <td>0.204655</td>\n",
       "      <td>NoEmp</td>\n",
       "      <td>-0.000311</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   dset_naics_holdout model LoanNr_ChkDgt  target   NAICS   enc_val variable  \\\n",
       "0                   1  menc    1000146010       0  445299  0.204655    NoEmp   \n",
       "1                   1  menc    1000734007       0  512110  0.204655    NoEmp   \n",
       "2                   1  menc    1000745006       0  541810  0.204655    NoEmp   \n",
       "\n",
       "      value  enc_feat  \n",
       "0  0.088108         0  \n",
       "1  0.044254         0  \n",
       "2 -0.000311         0  "
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "long_df.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "5aea96f0-ef2a-4de9-9f46-7140ce7dff78",
   "metadata": {},
   "outputs": [],
   "source": [
    "long_df.to_parquet(Path(setup.temp_path).joinpath('82_DATA_shap_long.parquet'))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "13577059-7714-4ae4-9f77-08d9def5b253",
   "metadata": {},
   "source": [
    "##### Long version, aggregated for encoded vs otherlong_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "776dd75c-1c60-4853-916d-e089389fb1fb",
   "metadata": {},
   "outputs": [],
   "source": [
    "long_df = pd.read_parquet(Path(setup.temp_path).joinpath('82_DATA_shap_long.parquet'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "6056ae3e-e069-478e-a20d-48c7db0d23a2",
   "metadata": {},
   "outputs": [],
   "source": [
    "long_enc_agg = long_df \\\n",
    "    .groupby(['LoanNr_ChkDgt', 'dset_naics_holdout', 'model', 'enc_feat']) \\\n",
    "    ['value'] \\\n",
    "    .agg('sum') \\\n",
    "    .reset_index()\n",
    "long_enc_agg['value_abs'] = np.abs(long_enc_agg['value'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "8563ef39-a16a-4554-a857-148f6e4a6cdb",
   "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>LoanNr_ChkDgt</th>\n",
       "      <th>dset_naics_holdout</th>\n",
       "      <th>model</th>\n",
       "      <th>enc_feat</th>\n",
       "      <th>value</th>\n",
       "      <th>value_abs</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1000014003</td>\n",
       "      <td>0</td>\n",
       "      <td>menc</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.967923</td>\n",
       "      <td>0.967923</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1000014003</td>\n",
       "      <td>0</td>\n",
       "      <td>menc</td>\n",
       "      <td>1</td>\n",
       "      <td>0.300321</td>\n",
       "      <td>0.300321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1000014003</td>\n",
       "      <td>0</td>\n",
       "      <td>naics</td>\n",
       "      <td>0</td>\n",
       "      <td>-1.000643</td>\n",
       "      <td>1.000643</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  LoanNr_ChkDgt  dset_naics_holdout  model  enc_feat     value  value_abs\n",
       "0    1000014003                   0   menc         0 -0.967923   0.967923\n",
       "1    1000014003                   0   menc         1  0.300321   0.300321\n",
       "2    1000014003                   0  naics         0 -1.000643   1.000643"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "long_enc_agg.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "37391f40-a498-4917-8526-0be80f823812",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8256972, 6)"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "long_enc_agg.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6e3ee168-98ba-4e37-adf3-7f99211301cd",
   "metadata": {},
   "source": [
    "## Bar Plots"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "60784136-e097-4fb4-a8d2-00d0426974ac",
   "metadata": {},
   "source": [
    "##### Importance (abs value)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "690e66fe-78f6-48e5-a0f2-26f75540a001",
   "metadata": {},
   "outputs": [],
   "source": [
    "imp_comp_df = long_enc_agg.groupby(['dset_naics_holdout', 'model', 'enc_feat']) \\\n",
    "    ['value_abs'] \\\n",
    "    .agg('mean') \\\n",
    "    .reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "f8baadd4-a5d9-4c8e-90cc-6c24d8e11067",
   "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>dset_naics_holdout</th>\n",
       "      <th>model</th>\n",
       "      <th>enc_feat</th>\n",
       "      <th>value_abs</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>menc</td>\n",
       "      <td>1</td>\n",
       "      <td>0.340816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>naics</td>\n",
       "      <td>1</td>\n",
       "      <td>0.340379</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0</td>\n",
       "      <td>dgi</td>\n",
       "      <td>1</td>\n",
       "      <td>0.315791</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1</td>\n",
       "      <td>menc</td>\n",
       "      <td>1</td>\n",
       "      <td>0.003326</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1</td>\n",
       "      <td>naics</td>\n",
       "      <td>1</td>\n",
       "      <td>0.033535</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1</td>\n",
       "      <td>dgi</td>\n",
       "      <td>1</td>\n",
       "      <td>0.031835</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    dset_naics_holdout  model  enc_feat  value_abs\n",
       "1                    0   menc         1   0.340816\n",
       "3                    0  naics         1   0.340379\n",
       "5                    0    dgi         1   0.315791\n",
       "7                    1   menc         1   0.003326\n",
       "9                    1  naics         1   0.033535\n",
       "11                   1    dgi         1   0.031835"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "imp_comp_plot_1 = imp_comp_df[imp_comp_df['enc_feat'] == 1]\n",
    "imp_comp_plot_1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "6e6373e0-234e-4941-b9fd-df14206b3a69",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(2, 1, sharex=True,\n",
    "                      layout='constrained')\n",
    "imp_comp_plot_1[imp_comp_plot_1['dset_naics_holdout'] == 0] \\\n",
    "    .plot(x='model', y='value_abs', ax=ax[0], kind='bar', legend=None)\n",
    "ax[0].yaxis.set_major_formatter(mtick.PercentFormatter(1, decimals=0))\n",
    "ax[0].set_title('test loans')\n",
    "imp_comp_plot_1[imp_comp_plot_1['dset_naics_holdout'] == 1] \\\n",
    "    .plot(x='model', y='value_abs', ax=ax[1], kind='bar', legend=None)\n",
    "ax[1].set_title('holdout NAICS loans')\n",
    "ax[1].set_xticklabels(['target enc', 'hier NAICS', 'hier DGI'], rotation=0)\n",
    "ax[1].set_xlabel(None)\n",
    "ax[1].yaxis.set_major_formatter(mtick.PercentFormatter(1, decimals=0))\n",
    "fig.supylabel('Encoding SHAP importance')\n",
    "fig.savefig(Path(setup.temp_path).joinpath('82_PLOT_importance_feat.png'))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7661fc4f-feda-4cb1-be68-20af81cd8f1b",
   "metadata": {},
   "source": [
    "##### Holdout means"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "ef4c4b29-da4b-435d-ae38-3166a7810559",
   "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>model</th>\n",
       "      <th>value</th>\n",
       "      <th>enc_val</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>menc</td>\n",
       "      <td>0.018266</td>\n",
       "      <td>0.204655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>naics</td>\n",
       "      <td>-0.064118</td>\n",
       "      <td>0.196476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>dgi</td>\n",
       "      <td>-0.176095</td>\n",
       "      <td>0.175107</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   model     value   enc_val\n",
       "0   menc  0.018266  0.204655\n",
       "1  naics -0.064118  0.196476\n",
       "2    dgi -0.176095  0.175107"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ho_comp_df_1 = long_df[(long_df['enc_feat'] == 1) &\n",
    "                             (long_df['dset_naics_holdout'] == 1)] \\\n",
    "    .groupby('model') \\\n",
    "    [['value', 'enc_val']] \\\n",
    "    .agg('mean') \\\n",
    "    .reset_index()\n",
    "ho_comp_df_1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "6551b0ff-3e5e-4529-8276-486e173d8dbc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(2, 1, sharex=True,\n",
    "                      layout='constrained')\n",
    "ho_comp_df_1.plot(x='model', y='enc_val', ax=ax[0], kind='bar', legend=None)\n",
    "ax[0].yaxis.set_major_formatter(mtick.PercentFormatter(1, decimals=0))\n",
    "ax[0].set_ylabel('mean encoded value')\n",
    "ho_comp_df_1.plot(x='model', y='value', ax=ax[1], kind='bar', legend=None)\n",
    "ax[1].set_ylabel('SHAP value')\n",
    "ax[1].set_xticklabels(['target enc', 'hier NAICS', 'hier DGI'], rotation=0)\n",
    "ax[1].set_xlabel(None)\n",
    "[xmin, xmax] =ax[1].get_xlim()\n",
    "ax[1].hlines(y=0, xmin=xmin, xmax=xmax, color='darkgray')\n",
    "fig.suptitle('Holdout Data Encoding Values')\n",
    "fig.savefig(Path(setup.temp_path).joinpath('82_PLOT_encoding_values_shap_holdout.png'))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b0ea9509-6100-446e-9a7f-2fc74c38514c",
   "metadata": {},
   "source": [
    "## Value Histograms"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d528966e-f6aa-491a-b0ca-078a1bb10c3e",
   "metadata": {},
   "source": [
    "##### Holdout data, SHAP values for the encoding feature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "2624986f-5dd7-4b27-a354-ba97612f4778",
   "metadata": {},
   "outputs": [],
   "source": [
    "val_hist_plot_1 = long_df[(long_df['enc_feat'] == 1) & \n",
    "                         (long_df['dset_naics_holdout'] == 1)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "88dbff6b-b412-49c6-a63c-9d78817cbab8",
   "metadata": {},
   "outputs": [],
   "source": [
    "groups = val_hist_plot_1.groupby('model')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "88fed452-c850-486f-8bf6-2a9729959ecd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(groups)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "629bdde9-cd68-4610-8cd5-c17f30738f22",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_lab = ['target encoded NAICS', 'hier enc by NAICS', 'hier enc by DGI']\n",
    "fig, ax = plt.subplots(num_models, 1, figsize=(6,4), sharex= True, sharey = True,\n",
    "                      layout='constrained')\n",
    "i=0\n",
    "for n, g in groups:\n",
    "    g['value'].hist(ax=ax[i], bins=40)\n",
    "    ax[i].set_title(plot_lab[i])\n",
    "    i+=1\n",
    "ax[i-1].set_xlabel('SHAP value for encoding')\n",
    "ax[i-1].set_xlim(-1, 1, auto=False)\n",
    "fig.suptitle('Holdout NAICS')\n",
    "fig.savefig(Path(setup.temp_path).joinpath('82_PLOT_histpgram_holdout_encoding_values.png'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "6368192f-0744-4ff7-b7d8-fd7ec851a718",
   "metadata": {},
   "outputs": [],
   "source": [
    "val_hist_plot_2 = long_df[(long_df['enc_feat'] == 1) & \n",
    "                         (long_df['dset_naics_holdout'] == 0)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "e4396e3c-1033-4d9d-9bed-f21b1297f6ec",
   "metadata": {},
   "outputs": [],
   "source": [
    "groups = val_hist_plot_2.groupby('model')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "d9185c9e-2d4e-4a92-af60-aadf6e1ecb9b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_lab = ['target encoded NAICS', 'hier enc by NAICS', 'hier enc by DGI']\n",
    "fig, ax = plt.subplots(num_models, 1, figsize=(6,4), sharex= True, sharey = True,\n",
    "                      layout='constrained')\n",
    "i=0\n",
    "for n, g in groups:\n",
    "    g['value'].hist(ax=ax[i], bins=40)\n",
    "    ax[i].set_title(plot_lab[i])\n",
    "    i+=1\n",
    "ax[i-1].set_xlabel('SHAP value for encoding')\n",
    "ax[i-1].set_xlim(-1, 1, auto=False)\n",
    "fig.suptitle('Test NAICS')\n",
    "fig.savefig(Path(setup.temp_path).joinpath('82_PLOT_histpgram_test_encoding_values.png'))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "54a16162-d404-45a3-bb76-b97033700cb8",
   "metadata": {},
   "source": [
    "## SHAP dependence"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "43902ae1-1611-4db8-b27f-a134171a1cbf",
   "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>dset_naics_holdout</th>\n",
       "      <th>model</th>\n",
       "      <th>LoanNr_ChkDgt</th>\n",
       "      <th>target</th>\n",
       "      <th>NAICS</th>\n",
       "      <th>enc_val</th>\n",
       "      <th>variable</th>\n",
       "      <th>value</th>\n",
       "      <th>enc_feat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>menc</td>\n",
       "      <td>1000146010</td>\n",
       "      <td>0</td>\n",
       "      <td>445299</td>\n",
       "      <td>0.204655</td>\n",
       "      <td>NoEmp</td>\n",
       "      <td>0.088108</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>menc</td>\n",
       "      <td>1000734007</td>\n",
       "      <td>0</td>\n",
       "      <td>512110</td>\n",
       "      <td>0.204655</td>\n",
       "      <td>NoEmp</td>\n",
       "      <td>0.044254</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>menc</td>\n",
       "      <td>1000745006</td>\n",
       "      <td>0</td>\n",
       "      <td>541810</td>\n",
       "      <td>0.204655</td>\n",
       "      <td>NoEmp</td>\n",
       "      <td>-0.000311</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   dset_naics_holdout model LoanNr_ChkDgt  target   NAICS   enc_val variable  \\\n",
       "0                   1  menc    1000146010       0  445299  0.204655    NoEmp   \n",
       "1                   1  menc    1000734007       0  512110  0.204655    NoEmp   \n",
       "2                   1  menc    1000745006       0  541810  0.204655    NoEmp   \n",
       "\n",
       "      value  enc_feat  \n",
       "0  0.088108         0  \n",
       "1  0.044254         0  \n",
       "2 -0.000311         0  "
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "long_df.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "d791e230-6cf8-4c21-ac28-5ba4dad13884",
   "metadata": {},
   "outputs": [],
   "source": [
    "dep_plot_1 = long_df[long_df['enc_feat'] == 1] \\\n",
    "    [['model', 'enc_val', 'value']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "6c361982-760a-4afc-9f25-672e6e99b140",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "groups = dep_plot_1.groupby('model')\n",
    "colors=['blue', 'darkorange', 'teal']\n",
    "labels = ['target enc', 'hier NAICS', 'hier DGI']\n",
    "i = 0\n",
    "for n, g in groups:\n",
    "    g.sample(500).plot(x='enc_val', y='value', kind='scatter', ax=ax, color=colors[i],\n",
    "                       label = labels[i])\n",
    "    i+=1\n",
    "ax.set_xlabel('feature value')\n",
    "ax.set_ylabel('SHAP value')\n",
    "fig.savefig(Path(setup.temp_path).joinpath('82_PLOT_shap_dependence.png'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "2fa7657f-992d-4883-8fce-d0824b64b55a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Smoothed dependence plot\n",
    "dep_plot_1['bin'] = pd.qcut(dep_plot_1['enc_val'], 20)\n",
    "dep_plot_2 = dep_plot_1.groupby(['model', 'bin']) \\\n",
    "    [['enc_val', 'value']] \\\n",
    "    .agg('mean') \\\n",
    "    .reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "0db3c77e-b8e3-44af-8748-38bf3fd004d7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "groups = dep_plot_2.groupby('model')\n",
    "colors=['blue', 'darkorange', 'teal']\n",
    "labels = ['target encoding', 'hierarchical NAICS', 'hierchical DGI']\n",
    "i = 0\n",
    "for n, g in groups:\n",
    "    g.plot(x='enc_val', y='value', kind='line', ax=ax, color=colors[i],\n",
    "                       label = labels[i])\n",
    "    i+=1\n",
    "ax.set_xlabel('feature value')\n",
    "ax.set_ylabel('SHAP value')\n",
    "ax.set_title('SHAP dependence (smoothed)')\n",
    "fig.savefig(Path(setup.temp_path).joinpath('82_PLOT_shap_dependence_smoothed.png'),\n",
    "           bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b4d76a7a-1115-435e-9ff5-ff40c9172821",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "tf_p39",
   "language": "python",
   "name": "tf_p39"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}