81_groupings_pred_err.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "70ef275c-a585-4b20-a706-75e58a10afe2",
   "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": "521fcd5f-0d5c-46f0-bcbd-8de97731f7e1",
   "metadata": {},
   "source": [
    "# Embedding Feature Errors, Performance\n",
    "In hierarchical encoding, more general NAICS categories are worse than specific.  For DGI clusters, it's the opposite.  Why? "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "285dade8-6474-4a7d-a7be-a37f39badd82",
   "metadata": {},
   "source": [
    "This notebook contains some examination of error rates for NAICS vs DGI encodings.  It also looks at the performance of encodings alone"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "fa66ce37-7440-40be-977a-76cd1be13138",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from pathlib import Path\n",
    "import importlib, os, pickle, re"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "4487fc7c-3404-40ea-ae81-6e5de75ccc9f",
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "1a58f621-8f48-4369-9aef-44915ccb99b0",
   "metadata": {},
   "outputs": [],
   "source": [
    "import sklearn as sk"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "590ef7d4-1aba-4ce1-a3b4-859649c448bb",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sba_gnn.sba_gnn import sg_plot \n",
    "sg_plot.plot_defaults()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "3d0d0273-12d4-4610-9e6f-ca020dbe3c58",
   "metadata": {},
   "outputs": [],
   "source": [
    "import shap\n",
    "import xgboost as xgb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "147c225f-4f9c-4aae-8a9b-528ed079621e",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "08b1ed43-049a-4683-992a-b743f07615c6",
   "metadata": {},
   "source": [
    "## Import Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "a3afb8ea-fd53-47be-93f3-7c090ef925b6",
   "metadata": {},
   "outputs": [],
   "source": [
    "sba_loans = pd.read_parquet(Path(setup.temp_path).joinpath('01_DATA_transformed.parquet'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "b5e04795-e35d-45ea-8627-e60a01f02625",
   "metadata": {},
   "outputs": [],
   "source": [
    "embed_df = pd.read_parquet(Path(setup.temp_path).joinpath('63_DATA_embeddings_tsne_naics.parquet'))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "11f9249a-b9f1-4085-a16e-9d6be1ee62e0",
   "metadata": {
    "tags": []
   },
   "source": [
    "##### Hierarchical encodings only\n",
    "Map the DGI based encoded feature into the loans"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "3320a069-32e6-42d3-bd20-e5dcb5839117",
   "metadata": {},
   "outputs": [],
   "source": [
    "hier_df = pd.read_parquet(Path(setup.temp_path).joinpath('65_DATA_dgi_mhier.parquet'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "d538111b-74bb-4722-9aef-e190fd0f55ef",
   "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": 12,
   "id": "e6017b3b-1d26-4cd8-9846-0043a47271e7",
   "metadata": {},
   "outputs": [],
   "source": [
    "sba_loans = sba_loans.merge(hier_df[['LoanNr_ChkDgt','menc_dgi']],\n",
    "                            on='LoanNr_ChkDgt')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "8c496798-912e-4baf-afe8-f892cdb44957",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(688081, 59)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sba_loans.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dbde81bd-8771-4ae8-bc30-8aa4480861fd",
   "metadata": {},
   "source": [
    "##### Get predictors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "d17f12ac-c2dc-4ff8-ba3a-ac4cfdb66fb5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['NoEmp', 'CreateJob', 'LowDoc', 'DisbursementGross', 'new_business', 'urban_flag', 'franchise_flag', 'mhier_NAICS']\n"
     ]
    }
   ],
   "source": [
    "with open(Path(setup.temp_path).joinpath('10_DATA_features_predict.pkl'), 'rb') as fin:\n",
    "    predictors_naics = pickle.load(fin)\n",
    "print(predictors_naics)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ad660472-e15e-4098-a20b-5718ce8e16b5",
   "metadata": {},
   "source": [
    "##### Predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "350c6227-e4fa-4671-af85-8f1b4ed37e07",
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_dgi = pd.read_parquet(Path(setup.temp_path).joinpath('65_DATA_combined_predictions.parquet'))\n",
    "pred_naics = pd.read_parquet(Path(setup.temp_path).joinpath('10_DATA_combined_predictions.parquet'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "e94e757f-f4d1-4d9b-b3c2-5683992d210b",
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_dgi = pred_dgi.merge(sba_loans[['LoanNr_ChkDgt', 'NAICS', 'menc_dgi', 'menc_NAICS']],\n",
    "                          on='LoanNr_ChkDgt')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "7740a36e-b579-4098-92e8-a2839f2378a8",
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_naics = pred_naics.drop(columns='NAICS', errors='ignore') \\\n",
    "    .merge(sba_loans[['LoanNr_ChkDgt', 'NAICS', 'mhier_NAICS', 'menc_NAICS']],\n",
    "                          on='LoanNr_ChkDgt')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cfb5ee6a-3962-4db2-addd-971a928ee68f",
   "metadata": {},
   "source": [
    "## Encoded Values - Metrics\n",
    "Look at the performance of the encodings alone, separate from XGBoost or other models. What is the performance on train, test, holdout? "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "73f0f775-b9b5-4946-be29-71b1239c2767",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Do a simple threshold to get basic metrics; this is for comparison \n",
    "pred_naics['pred_bin_targ'] = np.where(pred_naics['menc_NAICS'] >= 0.2, 1, 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "8d483fe6-a542-4252-a007-4c72ea1a3904",
   "metadata": {},
   "outputs": [],
   "source": [
    "# How good is the standard target encoding as its own model?\n",
    "metrics_mhier_targ_dset  = pred_naics.groupby('dset') \\\n",
    "    .apply(lambda x: sg_plot.dset_metrics(x.target, x.pred_bin_targ, x.pred_bin_targ)) \\\n",
    "    .reset_index()\n",
    "metrics_mhier_targ_dset_test = pred_naics[pred_naics['dset'] == 'test'] \\\n",
    "    .groupby(['dset', 'dset_naics_holdout']) \\\n",
    "    .apply(lambda x: sg_plot.dset_metrics(x.target, x.pred_bin_targ, x.pred_bin_targ)) \\\n",
    "    .reset_index()\n",
    "metrics_mhier_targ = pd.concat([metrics_mhier_targ_dset, metrics_mhier_targ_dset_test])\n",
    "metrics_mhier_targ\n",
    "metrics_mhier_targ.to_csv(Path(setup.temp_path).joinpath('81_REPORT_embedding_perf_menc_NAICS.csv'), index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "d911e59e-0e6e-4ef5-9fb4-800beab38e0c",
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_naics['pred_bin_menc'] = np.where(pred_naics['mhier_NAICS'] >= 0.2, 1, 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "153b003c-41df-4a39-9849-70367ab6448d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# What about hiererchical encoding by NAICS?\n",
    "metrics_mhier_naics_dset  = pred_naics.groupby('dset') \\\n",
    "    .apply(lambda x: sg_plot.dset_metrics(x.target, x.pred_bin_menc, x.mhier_NAICS)) \\\n",
    "    .reset_index()\n",
    "metrics_mhier_naics_dset_test = pred_naics[pred_naics['dset'] == 'test'] \\\n",
    "    .groupby(['dset', 'dset_naics_holdout']) \\\n",
    "    .apply(lambda x: sg_plot.dset_metrics(x.target, x.pred_bin_menc, x.mhier_NAICS)) \\\n",
    "    .reset_index()\n",
    "metrics_mhier_naics = pd.concat([metrics_mhier_naics_dset, metrics_mhier_naics_dset_test])\n",
    "metrics_mhier_naics\n",
    "metrics_mhier_naics.to_csv(Path(setup.temp_path).joinpath('81_REPORT_embedding_perf_mhier_NAICS.csv'), index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "c5ad439a-1aed-450b-9fc2-50982621727a",
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_dgi['pred_bin_menc'] = np.where(pred_dgi['menc_dgi'] >= 0.2, 1, 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "e01b7a57-100c-4413-84f4-ee07ca4d3260",
   "metadata": {},
   "outputs": [],
   "source": [
    "# What about DGI encoding\n",
    "metrics_mhier_dgi_dset  = pred_dgi.groupby('dset') \\\n",
    "    .apply(lambda x: sg_plot.dset_metrics(x.target, x.pred_bin_menc, x.menc_dgi)) \\\n",
    "    .reset_index()\n",
    "metrics_mhier_dgi_dset_test = pred_dgi[pred_dgi['dset'] == 'test'] \\\n",
    "    .groupby(['dset', 'dset_naics_holdout']) \\\n",
    "    .apply(lambda x: sg_plot.dset_metrics(x.target, x.pred_bin_menc, x.menc_dgi)) \\\n",
    "    .reset_index()\n",
    "metrics_mhier_dgi = pd.concat([metrics_mhier_dgi_dset, metrics_mhier_dgi_dset_test])\n",
    "metrics_mhier_dgi\n",
    "metrics_mhier_dgi.to_csv(Path(setup.temp_path).joinpath('81_REPORT_embedding_perf_mehc_dgi.csv'), index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "24c66ff2-f7ab-4c9b-b4f8-d11940ff01a0",
   "metadata": {},
   "source": [
    "## Histograms - Encoded Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "b1166927-b7c6-42af-b1ce-f40b129cec11",
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_dgi_ho = pred_dgi[pred_dgi['dset_naics_holdout'] ==1].copy()\n",
    "pred_naics_ho = pred_naics[pred_naics['dset_naics_holdout'] ==1].copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "ac0837a8-7839-46c5-979d-b4136a70021d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    688081.000000\n",
       "mean          0.203906\n",
       "std           0.095690\n",
       "min           0.012987\n",
       "25%           0.126003\n",
       "50%           0.205592\n",
       "75%           0.264286\n",
       "max           0.572148\n",
       "Name: mhier_NAICS, dtype: float64"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_naics['mhier_NAICS'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "1553594d-f658-41b7-9642-ffd5a19f735e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    65052.000000\n",
       "mean         0.196476\n",
       "std          0.082249\n",
       "min          0.036665\n",
       "25%          0.110215\n",
       "50%          0.226657\n",
       "75%          0.254464\n",
       "max          0.351724\n",
       "Name: mhier_NAICS, dtype: float64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_naics_ho['mhier_NAICS'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "e04219f3-58b3-4f40-81a5-017507c64870",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    688081.000000\n",
       "mean          0.201907\n",
       "std           0.093596\n",
       "min           0.006059\n",
       "25%           0.126132\n",
       "50%           0.188385\n",
       "75%           0.260897\n",
       "max           0.642367\n",
       "Name: menc_dgi, dtype: float64"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_dgi['menc_dgi'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "7060fd7d-abc5-4ded-80a1-4db7bf8e9190",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0.98, 'Holdout Encoded Values')"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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, sharey=True,\n",
    "                       layout='constrained')\n",
    "pred_naics_ho['mhier_NAICS'].hist(ax=ax[0])\n",
    "ax[0].set_title('NAICS hierarchy', size=12)\n",
    "pred_dgi_ho['menc_dgi'].hist(ax=ax[1])\n",
    "ax[1].set_title('DGI hierarchy', size=12)\n",
    "#ax[1].set_xlim([0,0.5])\n",
    "fig.supylabel('count')\n",
    "fig.supxlabel('Encoded Value')\n",
    "fig.suptitle('Holdout Encoded Values')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "4b0ad76b-79de-4679-b059-bcbfd56fa102",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0.98, 'All Encoded Values')"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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, sharey=True,\n",
    "                       layout='constrained')\n",
    "pred_naics['mhier_NAICS'].hist(ax=ax[0])\n",
    "ax[0].set_title('NAICS hierarchy', size=12)\n",
    "pred_dgi['menc_dgi'].hist(ax=ax[1])\n",
    "ax[1].set_title('DGI hierarchy', size=12)\n",
    "#ax[1].set_xlim([0,0.5])\n",
    "fig.supylabel('count')\n",
    "fig.supxlabel('Encoded Value')\n",
    "fig.suptitle('All Encoded Values')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f1379d9e-051f-4f01-8298-02fbba52db80",
   "metadata": {},
   "source": [
    "## Histograms - By NAICS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "d3fda2a9-b563-48a8-8579-439c1d22c6fe",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to get error, other information by NAICS\n",
    "def get_err_by_naics(pred_df, hier_feat = 'menc_dgi'):\n",
    "    ho_sum = pred_df \\\n",
    "        .groupby(['NAICS', hier_feat]) \\\n",
    "        .agg({'LoanNr_ChkDgt':'count', 'target':['sum', 'mean'], 'predict_prob':'mean',\n",
    "              'predict_bin':'sum'}) \\\n",
    "    .reset_index()  \n",
    "    ho_sum.columns = [re.sub('_$', '', '_'.join(c)) for c in ho_sum.columns]\n",
    "    ho_sum.rename(columns={'LoanNr_ChkDgt_count':'count'}, inplace=True)\n",
    "    ho_sum['prob_err'] = ho_sum['predict_prob_mean'] - ho_sum['target_mean']\n",
    "    return ho_sum"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "a10e8175-a33e-4a22-85c6-bdceb9ba6bdc",
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_dgi_grp = get_err_by_naics(pred_dgi)\n",
    "pred_naics_grp = get_err_by_naics(pred_naics, 'mhier_NAICS')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "270d56e3-ceaa-4887-b75f-d6f8df94ae58",
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_dgi_grp_ho = get_err_by_naics(pred_dgi_ho)\n",
    "pred_naics_grp_ho = get_err_by_naics(pred_naics_ho, 'mhier_NAICS')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "e0f8869a-95ba-4527-80ff-e7c7f5419507",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0.98, 'Holdout Encoded Values (NAICS level)')"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Histogram- holdout by NAICS\n",
    "fig, ax = plt.subplots(3,1, sharex=True, sharey=False,\n",
    "                       layout='constrained')\n",
    "pred_naics_grp_ho['target_mean'].hist(ax=ax[0])\n",
    "ax[0].set_title('Actual Target Rate', size=12)\n",
    "pred_naics_grp_ho['mhier_NAICS'].hist(ax=ax[1])\n",
    "ax[1].set_title('NAICS hierarchy', size=12)\n",
    "pred_dgi_grp_ho['menc_dgi'].hist(ax=ax[2])\n",
    "ax[2].set_title('DGI hierarchy', size=12)\n",
    "#ax[2].set_xlim([0,0.5])\n",
    "fig.supylabel('count')\n",
    "fig.supxlabel('Encoded Value')\n",
    "fig.suptitle('Holdout Encoded Values (NAICS level)')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "18e8eefd-4b16-4637-804b-2e8cb78066ad",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0.98, 'All Encoded Values')"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(3,1, sharex=True, sharey=False,\n",
    "                       layout='constrained')\n",
    "pred_naics_grp[pred_naics_grp['count'] >= 100]['target_mean'].hist(ax=ax[0])\n",
    "ax[0].set_title('Actual Target Rate (>150 counts)', size=12)\n",
    "pred_naics_grp['mhier_NAICS'].hist(ax=ax[1])\n",
    "ax[1].set_title('NAICS hierarchy', size=12)\n",
    "pred_dgi_grp['menc_dgi'].hist(ax=ax[2])\n",
    "ax[2].set_title('DGI hierarchy', size=12)\n",
    "ax[2].set_xlim([0,0.6])\n",
    "fig.supylabel('count')\n",
    "fig.supxlabel('Encoded Value')\n",
    "fig.suptitle('All Encoded Values')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "499d19d3-1340-411a-866b-1215b06b1d25",
   "metadata": {},
   "source": [
    "## Holdout - error analysis by NAICS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "8e018709-cf10-42ef-bc24-a2a2060d6aad",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='menc_dgi', ylabel='prob_err'>"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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",
    "pred_naics_grp_ho[pred_naics_grp_ho['count'] >= 100].plot(x='mhier_NAICS', y='prob_err', kind='scatter', ax=ax)\n",
    "pred_dgi_grp_ho[pred_dgi_grp_ho['count'] >= 100] \\\n",
    "    .plot(x='menc_dgi', y='prob_err', kind='scatter', ax = ax, color='darkorange')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8ebf66a2-dac5-43dc-b70a-39e293eff8c4",
   "metadata": {},
   "source": [
    "## Sub-metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "e549b2b2-536f-44e6-9001-0b9d09b599a1",
   "metadata": {},
   "outputs": [],
   "source": [
    "bins = [-np.Inf, 0.2, np.Inf]\n",
    "pred_dgi_ho['pred_cut'] = pd.cut(pred_dgi_ho['menc_dgi'], bins=bins)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "8b7a1109-dbea-45c5-87df-cd41aed34a51",
   "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>pred_cut</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>0</th>\n",
       "      <td>(-inf, 0.2]</td>\n",
       "      <td>0.800967</td>\n",
       "      <td>0.280176</td>\n",
       "      <td>0.301868</td>\n",
       "      <td>0.261394</td>\n",
       "      <td>0.285785</td>\n",
       "      <td>0.715709</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>(0.2, inf]</td>\n",
       "      <td>0.487676</td>\n",
       "      <td>0.468094</td>\n",
       "      <td>0.321540</td>\n",
       "      <td>0.860132</td>\n",
       "      <td>0.374741</td>\n",
       "      <td>0.655205</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      pred_cut  accuracy_score  f1_score  precision_score  recall_score  \\\n",
       "0  (-inf, 0.2]        0.800967  0.280176         0.301868      0.261394   \n",
       "1   (0.2, inf]        0.487676  0.468094         0.321540      0.860132   \n",
       "\n",
       "   average_precision_score  roc_auc_score  \n",
       "0                 0.285785       0.715709  \n",
       "1                 0.374741       0.655205  "
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metrics_dgi_sub = pred_dgi_ho.groupby('pred_cut') \\\n",
    "    .apply(lambda x: sg_plot.dset_metrics(x.target, x.predict_bin, x.predict_prob)) \\\n",
    "    .reset_index()\n",
    "metrics_dgi_sub"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "9a5ef82e-4517-4de4-ae47-7dd7226ef808",
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_naics_ho['pred_cut'] = pd.cut(pred_naics_ho['mhier_NAICS'],  bins=bins)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "c23fa1a9-46a8-4ae5-9d5e-444fb056c1d1",
   "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>pred_cut</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>0</th>\n",
       "      <td>(-inf, 0.2]</td>\n",
       "      <td>0.831629</td>\n",
       "      <td>0.256321</td>\n",
       "      <td>0.297090</td>\n",
       "      <td>0.225391</td>\n",
       "      <td>0.247685</td>\n",
       "      <td>0.693689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>(0.2, inf]</td>\n",
       "      <td>0.606782</td>\n",
       "      <td>0.462433</td>\n",
       "      <td>0.327787</td>\n",
       "      <td>0.784813</td>\n",
       "      <td>0.402475</td>\n",
       "      <td>0.733532</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      pred_cut  accuracy_score  f1_score  precision_score  recall_score  \\\n",
       "0  (-inf, 0.2]        0.831629  0.256321         0.297090      0.225391   \n",
       "1   (0.2, inf]        0.606782  0.462433         0.327787      0.784813   \n",
       "\n",
       "   average_precision_score  roc_auc_score  \n",
       "0                 0.247685       0.693689  \n",
       "1                 0.402475       0.733532  "
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metrics_naics_sub = pred_naics_ho.groupby('pred_cut') \\\n",
    "    .apply(lambda x: sg_plot.dset_metrics(x.target, x.predict_bin, x.predict_prob)) \\\n",
    "    .reset_index()\n",
    "metrics_naics_sub"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e56c91a4-b4f9-493b-a87c-41a42df2453d",
   "metadata": {},
   "source": [
    "##### Overall test data perf\n",
    "Random selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "9dd46547-32f7-4e9d-8fe3-588cd8802735",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "pred_naics_test = pred_naics[pred_naics['dset_naics_holdout'] ==  0].copy()\n",
    "pred_dgi_test = pred_dgi[pred_dgi['dset_naics_holdout'] ==  0].copy()\n",
    "pred_dgi_test['pred_cut'] = pd.cut(pred_dgi_test['menc_dgi'], bins=bins)\n",
    "pred_naics_test['pred_cut'] = pd.cut(pred_naics_test['mhier_NAICS'],  bins=bins)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "ee803182-bd75-498f-ba5f-bad38ac303a3",
   "metadata": {
    "tags": []
   },
   "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>pred_cut</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>0</th>\n",
       "      <td>(-inf, 0.2]</td>\n",
       "      <td>0.819585</td>\n",
       "      <td>0.313959</td>\n",
       "      <td>0.298265</td>\n",
       "      <td>0.331396</td>\n",
       "      <td>0.272064</td>\n",
       "      <td>0.736541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>(0.2, inf]</td>\n",
       "      <td>0.513683</td>\n",
       "      <td>0.490681</td>\n",
       "      <td>0.348543</td>\n",
       "      <td>0.828584</td>\n",
       "      <td>0.424507</td>\n",
       "      <td>0.668589</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      pred_cut  accuracy_score  f1_score  precision_score  recall_score  \\\n",
       "0  (-inf, 0.2]        0.819585  0.313959         0.298265      0.331396   \n",
       "1   (0.2, inf]        0.513683  0.490681         0.348543      0.828584   \n",
       "\n",
       "   average_precision_score  roc_auc_score  \n",
       "0                 0.272064       0.736541  \n",
       "1                 0.424507       0.668589  "
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_naics_test.groupby('pred_cut') \\\n",
    "    .apply(lambda x: sg_plot.dset_metrics(x.target, x.predict_bin, x.predict_prob)) \\\n",
    "    .reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "ea72ece6-cf98-4237-985f-2f4d26235df1",
   "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>pred_cut</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>0</th>\n",
       "      <td>(-inf, 0.2]</td>\n",
       "      <td>0.813214</td>\n",
       "      <td>0.317851</td>\n",
       "      <td>0.296513</td>\n",
       "      <td>0.342498</td>\n",
       "      <td>0.273817</td>\n",
       "      <td>0.735094</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>(0.2, inf]</td>\n",
       "      <td>0.509217</td>\n",
       "      <td>0.488133</td>\n",
       "      <td>0.345525</td>\n",
       "      <td>0.831185</td>\n",
       "      <td>0.419338</td>\n",
       "      <td>0.666791</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      pred_cut  accuracy_score  f1_score  precision_score  recall_score  \\\n",
       "0  (-inf, 0.2]        0.813214  0.317851         0.296513      0.342498   \n",
       "1   (0.2, inf]        0.509217  0.488133         0.345525      0.831185   \n",
       "\n",
       "   average_precision_score  roc_auc_score  \n",
       "0                 0.273817       0.735094  \n",
       "1                 0.419338       0.666791  "
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_dgi_test.groupby('pred_cut') \\\n",
    "    .apply(lambda x: sg_plot.dset_metrics(x.target, x.predict_bin, x.predict_prob)) \\\n",
    "    .reset_index()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "47e4aefc-ec80-4312-88a4-6e671aa37d45",
   "metadata": {},
   "source": [
    "##### Training performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "14e0295a-e10c-480c-9790-6454af4dc20a",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "pred_naics_train = pred_naics[pred_naics['dset'] ==  'train'].copy()\n",
    "pred_dgi_train = pred_dgi[pred_dgi['dset'] ==  'train'].copy()\n",
    "pred_dgi_train['pred_cut'] = pd.cut(pred_dgi_train['menc_dgi'], bins=bins)\n",
    "pred_naics_train['pred_cut'] = pd.cut(pred_naics_train['mhier_NAICS'],  bins=bins)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "22e0def1-6bf3-4730-80f1-ddb40c6ee582",
   "metadata": {
    "tags": []
   },
   "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>pred_cut</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>0</th>\n",
       "      <td>(-inf, 0.2]</td>\n",
       "      <td>0.820685</td>\n",
       "      <td>0.313185</td>\n",
       "      <td>0.295814</td>\n",
       "      <td>0.332724</td>\n",
       "      <td>0.271698</td>\n",
       "      <td>0.738521</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>(0.2, inf]</td>\n",
       "      <td>0.514646</td>\n",
       "      <td>0.492109</td>\n",
       "      <td>0.349736</td>\n",
       "      <td>0.829986</td>\n",
       "      <td>0.427499</td>\n",
       "      <td>0.670211</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      pred_cut  accuracy_score  f1_score  precision_score  recall_score  \\\n",
       "0  (-inf, 0.2]        0.820685  0.313185         0.295814      0.332724   \n",
       "1   (0.2, inf]        0.514646  0.492109         0.349736      0.829986   \n",
       "\n",
       "   average_precision_score  roc_auc_score  \n",
       "0                 0.271698       0.738521  \n",
       "1                 0.427499       0.670211  "
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_naics_train.groupby('pred_cut') \\\n",
    "    .apply(lambda x: sg_plot.dset_metrics(x.target, x.predict_bin, x.predict_prob)) \\\n",
    "    .reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "52a692f6-520f-40d9-a644-0634e8ae9a1e",
   "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>pred_cut</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>0</th>\n",
       "      <td>(-inf, 0.2]</td>\n",
       "      <td>0.814443</td>\n",
       "      <td>0.317268</td>\n",
       "      <td>0.294545</td>\n",
       "      <td>0.343792</td>\n",
       "      <td>0.273254</td>\n",
       "      <td>0.736698</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>(0.2, inf]</td>\n",
       "      <td>0.510193</td>\n",
       "      <td>0.489541</td>\n",
       "      <td>0.346685</td>\n",
       "      <td>0.832642</td>\n",
       "      <td>0.422421</td>\n",
       "      <td>0.668602</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      pred_cut  accuracy_score  f1_score  precision_score  recall_score  \\\n",
       "0  (-inf, 0.2]        0.814443  0.317268         0.294545      0.343792   \n",
       "1   (0.2, inf]        0.510193  0.489541         0.346685      0.832642   \n",
       "\n",
       "   average_precision_score  roc_auc_score  \n",
       "0                 0.273254       0.736698  \n",
       "1                 0.422421       0.668602  "
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_dgi_train.groupby('pred_cut') \\\n",
    "    .apply(lambda x: sg_plot.dset_metrics(x.target, x.predict_bin, x.predict_prob)) \\\n",
    "    .reset_index()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b5b944f8-ba84-48df-910c-4a1eb5e41615",
   "metadata": {},
   "source": [
    "## Precision-Recall Curves"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "e19571cc-e862-4704-8654-5a8f3c052aa9",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_pr_curve(pred_df):\n",
    "    cd = sk.metrics.precision_recall_curve(pred_df['target'], pred_df['predict_prob'])\n",
    "    return pd.DataFrame({'precision':cd[0],\n",
    "                         'recall': cd[1]})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bff6c82d-b0ab-4550-b15b-3dbf944352fd",
   "metadata": {},
   "source": [
    "##### Compare overalls, holdouts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "ae262b2c-c71a-45b0-b67e-d804b51852f7",
   "metadata": {},
   "outputs": [],
   "source": [
    "pr_curve_dgi = get_pr_curve(pred_dgi[pred_dgi['dset'] == 'test'])\n",
    "pr_curve_dgi_ho = get_pr_curve(pred_dgi_ho)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "f1fcea94-08f6-4e18-bbdd-20cbf16ec2af",
   "metadata": {},
   "outputs": [],
   "source": [
    "pr_curve_naics = get_pr_curve(pred_naics[pred_naics['dset'] == 'test'])\n",
    "pr_curve_naics_ho = get_pr_curve(pred_naics_ho)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "66031948-63f8-4170-932b-ecfc9191d7d9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Holdout NAICS')"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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(1, 2, sharey=True)\n",
    "pr_curve_naics.plot(x='recall', y='precision', ax=ax[0], label='naics')\n",
    "pr_curve_dgi.plot(x='recall', y='precision', ax=ax[0],color='darkorange', label='dgi')\n",
    "ax[0].set_title('Test data')\n",
    "pr_curve_naics_ho.plot(x='recall', y='precision', ax=ax[1], label='naics')\n",
    "pr_curve_dgi_ho.plot(x='recall', y='precision', ax=ax[1], color='darkorange', label='dgi')\n",
    "ax[1].set_title('Holdout NAICS')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8ef5f468-c240-4961-b80a-53fc0c3908fc",
   "metadata": {},
   "source": [
    "##### Break Down Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "858ab9bf-7646-4900-8003-00c92cbd839c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#ig, ax = plt.subplots()\n",
    "b = [0, 0.2, np.Inf]\n",
    "for i in range(len(b)-1):\n",
    "    this_curve_naics = get_pr_curve(pred_naics_ho[(pred_naics_ho['mhier_NAICS'] >=b[i]) & \\\n",
    "                                   (pred_naics_ho['mhier_NAICS'] <=b[i+1])])\n",
    "    this_curve_dgi = get_pr_curve(pred_dgi_ho[(pred_dgi_ho['menc_dgi'] >=b[i]) & \\\n",
    "                                   (pred_dgi_ho['menc_dgi'] <=b[i+1])])\n",
    "    fig, ax = plt.subplots()\n",
    "    this_curve_naics.plot(x='recall', y='precision', ax=ax)\n",
    "    this_curve_dgi.plot(x='recall', y='precision', ax=ax,color='darkorange')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c71a223a-b30f-4ec6-b703-2e7f0a93ad5a",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dc3a064e-a86f-431c-a5ee-0b6bf263dc54",
   "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
}