44_random_cat.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "8816b366-5be4-4748-b769-4dc6f7d1735f",
   "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": "a945de0c-f042-46b2-9d06-ab064af6c05b",
   "metadata": {},
   "source": [
    "## Base model with single random categorical\n",
    "Raw model - add a random categorical with 1,000 levels.  No randomization.  No NAICS"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d47a7ec9-7d7b-47df-a902-b2f1e54f1135",
   "metadata": {},
   "source": [
    "*This script takes about 2 hours on my MacBook Air*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "fb42f9b4-99a8-4753-aacf-8722469be4c8",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import tensorflow as tf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "bf51170a-4487-4d10-b63d-c88f41b529f9",
   "metadata": {},
   "outputs": [],
   "source": [
    "from pathlib import Path\n",
    "import importlib, pickle\n",
    "from sklearn import model_selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "09dd973b-60a0-434b-9193-8c5fcec27dd5",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import os\n",
    "import re\n",
    "\n",
    "import keras\n",
    "from tensorflow.keras import layers, optimizers, losses, metrics, Model\n",
    "from sklearn import preprocessing, feature_extraction, model_selection\n",
    "from IPython.display import display, HTML\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "22b466e7-a419-45aa-83f7-5dafc7a3998b",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.layers import Dense, Dropout, Input, Embedding, \\\n",
    "    Concatenate, Reshape, concatenate, Flatten\n",
    "import tensorflow.keras.metrics as km"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "9f1e4f2e-0e45-4e28-8e9c-212e0bb51199",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sba_nn.sba_nn import sbnn_metrics, sbnn_model "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "b6cc1a60-d316-4c1b-8f61-c4bd0ec67363",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Imputer object for easy dataset conversion to NN friendly format\n",
    "from sba_nn.sba_nn import sbnn_imputer \n",
    "from sba_nn.sba_nn.sbnn_imputer import NNImputer"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3cbc00d2-bc83-42d9-83eb-65fcf306dccd",
   "metadata": {},
   "source": [
    "## Input Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "e014cd31-0f8d-42e3-82f9-b529eec8a3cc",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "sba_loans = pd.read_parquet(Path(setup.temp_path).joinpath('01_DATA_transformed_nomiss.parquet'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "8e65cb7c-792c-46ff-a530-0d5c64240d60",
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(Path(setup.temp_path).joinpath('01_DATA_features.pkl'), 'rb') as fin:\n",
    "    imputer_features = pickle.load(fin)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8dedec24-9c5f-4b53-b1d0-775119895ca1",
   "metadata": {},
   "source": [
    "## Modify Data\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "9594fc0c-8031-41f1-9ef2-f4e9b3581525",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Single random categorical\n",
    "sba_loans['rand_cat'] = pd.Series(np.random.randint(0,1000, size=len(sba_loans)),\n",
    "                                  index=sba_loans.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "ac4ca519-b360-4802-a48b-7604247a14e0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "999"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_len = sba_loans[sba_loans['dset'] == 'train']['rand_cat'].max()\n",
    "train_len"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "63fe3d0e-6dfc-43f8-bbf9-84a4bb023a8b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save data\n",
    "sba_loans[['LoanNr_ChkDgt', 'dset', 'dset_naics_holdout', 'NAICS', 'rand_cat']] \\\n",
    "     .to_parquet(Path(setup.temp_path).joinpath('44_DATA_rand_grp.parquet'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "babed9a5-b720-4012-9789-01410ecee616",
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(Path(setup.temp_path).joinpath('44_DATA_rand_max_encodings.pkl'), 'wb') as fout:\n",
    "    pickle.dump(train_len, fout)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4d4116da-d689-478a-9c5d-a330f231e615",
   "metadata": {},
   "source": [
    "# Features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "f98c6e41-38d0-4d5f-b821-e3837b8c2068",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['NoEmp', 'CreateJob', 'LowDoc', 'DisbursementGross', 'new_business', 'urban_flag', 'franchise_flag', 'missingindicator_LowDoc', 'missingindicator_new_business', 'missingindicator_urban_flag']\n"
     ]
    }
   ],
   "source": [
    "# Numeric features\n",
    "features_numeric = [f for f in imputer_features if 'NAICS' not in f]\n",
    "features_numeric_len = len(features_numeric)\n",
    "print(features_numeric)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "a74fa386-4011-45ee-bdb5-d671db9ffa91",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Use analagous embedding dimensions\n",
    "features_grp_all = ['rand_cat'] \n",
    "grp_emb_dim = [8]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b979396c-c822-4940-a292-6e07f1b54f58",
   "metadata": {},
   "source": [
    "##### Datasets for train, validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "43bca011-a59d-4496-9f03-87f4e350a253",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = sba_loans[['dset', 'LoanNr_ChkDgt'] + features_numeric].set_index('LoanNr_ChkDgt').sort_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "aa162d01-ed73-4ae3-832e-4aeea09b70f4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "training X: (446848, 10), y:(446848,)\n"
     ]
    }
   ],
   "source": [
    "X_train = X[X['dset'] == 'train'].drop(columns='dset')\n",
    "y_train = sba_loans[sba_loans['dset'] == 'train'].set_index('LoanNr_ChkDgt').sort_index()['target']\n",
    "print(f'training X: {X_train.shape}, y:{y_train.shape}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "b32b5fd3-6f6e-4aec-8e20-0e2b71a4814c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "val X: (95604, 10), y:(95604,)\n"
     ]
    }
   ],
   "source": [
    "X_val = X[X['dset'] == 'val'].drop(columns='dset')\n",
    "y_val = sba_loans[sba_loans['dset'] == 'val'].set_index('LoanNr_ChkDgt').sort_index()['target']\n",
    "print(f'val X: {X_val.shape}, y:{y_val.shape}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "01b8ff96-159b-4283-b5fb-c4b5dbd277d5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.20230592953308507\n"
     ]
    }
   ],
   "source": [
    "base_thresh = y_train.mean()\n",
    "print(base_thresh)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "1b8ce4c9-6907-4b70-982a-b92a52b3159d",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_hier = [sba_loans[['dset', 'LoanNr_ChkDgt', 'rand_cat']] \\\n",
    "              .set_index('LoanNr_ChkDgt').sort_index()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "a6f129e0-a4d0-4d93-9886-153c9803d403",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_hier_train = [Xn[Xn['dset'] == 'train'].drop(columns='dset') for Xn in X_hier]\n",
    "X_hier_val =  [Xn[Xn['dset'] == 'val'].drop(columns='dset') for Xn in X_hier]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3db4ea1c-77b2-4721-b106-1e66fc502139",
   "metadata": {},
   "source": [
    "## Create, fit model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "4d48ddfa-580c-4321-9917-0bd176b67d87",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-05-29 06:26:42.078551: I metal_plugin/src/device/metal_device.cc:1154] Metal device set to: Apple M1\n",
      "2024-05-29 06:26:42.078611: I metal_plugin/src/device/metal_device.cc:296] systemMemory: 16.00 GB\n",
      "2024-05-29 06:26:42.078628: I metal_plugin/src/device/metal_device.cc:313] maxCacheSize: 5.33 GB\n",
      "2024-05-29 06:26:42.078675: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:303] Could not identify NUMA node of platform GPU ID 0, defaulting to 0. Your kernel may not have been built with NUMA support.\n",
      "2024-05-29 06:26:42.078694: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:269] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 0 MB memory) -> physical PluggableDevice (device: 0, name: METAL, pci bus id: <undefined>)\n"
     ]
    }
   ],
   "source": [
    "this_model = sbnn_model.create_emb_model(features_numeric_len,\n",
    "                                         [train_len],\n",
    "                                         grp_emb_dim,\n",
    "                                         naics_embedding_names = ['rand_cat'],\n",
    "                                         hidden_size = setup.nn_layer_sizes,\n",
    "                                         activation='tanh',\n",
    "                                         lr=setup.nn_learning_rate,\n",
    "                                         opt_func = setup.nn_optimizer,\n",
    "                                         dropout = setup.nn_dropout)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "47036621-ef13-4957-997b-f8ef86016f7f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "__________________________________________________________________________________________________\n",
      " Layer (type)                Output Shape                 Param #   Connected to                  \n",
      "==================================================================================================\n",
      " input_naics_00 (InputLayer  [(None, 1)]                  0         []                            \n",
      " )                                                                                                \n",
      "                                                                                                  \n",
      " emb_naics_00 (Embedding)    (None, 1, 8)                 7992      ['input_naics_00[0][0]']      \n",
      "                                                                                                  \n",
      " rand_cat (Flatten)          (None, 8)                    0         ['emb_naics_00[0][0]']        \n",
      "                                                                                                  \n",
      " input_numeric (InputLayer)  [(None, 10)]                 0         []                            \n",
      "                                                                                                  \n",
      " input_concat (Concatenate)  (None, 18)                   0         ['rand_cat[0][0]',            \n",
      "                                                                     'input_numeric[0][0]']       \n",
      "                                                                                                  \n",
      " layer_00 (Dense)            (None, 128)                  2432      ['input_concat[0][0]']        \n",
      "                                                                                                  \n",
      " dropout_00 (Dropout)        (None, 128)                  0         ['layer_00[0][0]']            \n",
      "                                                                                                  \n",
      " layer_01 (Dense)            (None, 64)                   8256      ['dropout_00[0][0]']          \n",
      "                                                                                                  \n",
      " dropout_01 (Dropout)        (None, 64)                   0         ['layer_01[0][0]']            \n",
      "                                                                                                  \n",
      " output (Dense)              (None, 1)                    65        ['dropout_01[0][0]']          \n",
      "                                                                                                  \n",
      "==================================================================================================\n",
      "Total params: 18745 (73.22 KB)\n",
      "Trainable params: 18745 (73.22 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "this_model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "cbb85f7c-127b-4a78-bcd0-bc00cc6bed14",
   "metadata": {},
   "outputs": [],
   "source": [
    "callback = keras.callbacks.EarlyStopping(patience=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "31772016-3643-468b-8d65-c456168cad6e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-05-29 06:26:56.315487: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "13963/13964 [============================>.] - ETA: 0s - loss: 0.4720 - auc: 0.3050 - auc_1: 0.6728"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-05-29 06:30:19.080028: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "13964/13964 [==============================] - 225s 16ms/step - loss: 0.4720 - auc: 0.3050 - auc_1: 0.6728 - val_loss: 0.4652 - val_auc: 0.3311 - val_auc_1: 0.6929\n",
      "Epoch 2/20\n",
      "13964/13964 [==============================] - 184s 13ms/step - loss: 0.4656 - auc: 0.3230 - auc_1: 0.6873 - val_loss: 0.4658 - val_auc: 0.3373 - val_auc_1: 0.6983\n",
      "Epoch 3/20\n",
      "13964/13964 [==============================] - 192s 14ms/step - loss: 0.4640 - auc: 0.3302 - auc_1: 0.6917 - val_loss: 0.4638 - val_auc: 0.3374 - val_auc_1: 0.6987\n",
      "Epoch 4/20\n",
      "13964/13964 [==============================] - 194s 14ms/step - loss: 0.4629 - auc: 0.3335 - auc_1: 0.6946 - val_loss: 0.4641 - val_auc: 0.3355 - val_auc_1: 0.6973\n",
      "Epoch 5/20\n",
      "13964/13964 [==============================] - 198s 14ms/step - loss: 0.4621 - auc: 0.3368 - auc_1: 0.6969 - val_loss: 0.4651 - val_auc: 0.3301 - val_auc_1: 0.6944\n",
      "Epoch 6/20\n",
      "13964/13964 [==============================] - 198s 14ms/step - loss: 0.4616 - auc: 0.3389 - auc_1: 0.6984 - val_loss: 0.4646 - val_auc: 0.3321 - val_auc_1: 0.6951\n"
     ]
    }
   ],
   "source": [
    "this_history = this_model.fit([X_train] + X_hier_train, y_train,\n",
    "                              validation_data=([X_val] + X_hier_val, y_val),\n",
    "                              batch_size=setup.nn_batch_size,\n",
    "                              epochs=setup.nn_epochs,\n",
    "                              callbacks=[callback])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "460c4b43-c673-4ee1-b070-d2d596ac4561",
   "metadata": {},
   "outputs": [],
   "source": [
    "this_history_df = sbnn_model.process_history(this_history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "ad4783f2-8d0b-4850-a45a-3a4be8e63318",
   "metadata": {},
   "outputs": [],
   "source": [
    "this_history_df.to_csv(Path(setup.temp_path).joinpath('44_REPORT_fit_history.csv'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "73e89efd-75a9-4b27-81a8-13e12d5482be",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 33,
     "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": [
    "this_history_df[['loss', 'val_loss']].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "2342a31f-4891-4834-851a-b90b8b729c3c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 34,
     "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": [
    "this_history_df[['auc', 'val_auc']].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "cee80637-cabf-4974-b476-08c5fba165ff",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "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": [
    "this_history_df[['auc_roc', 'val_auc_roc']].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "e184b0ca-1d31-4798-88a0-4e590b67062e",
   "metadata": {},
   "outputs": [],
   "source": [
    "this_model.save(Path(setup.temp_path).joinpath('44_DATA_model.keras'),save_format='tf')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d326c9ea-4759-4418-a64d-56c151049c25",
   "metadata": {},
   "source": [
    "## Predictions on all data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "d5c39a5a-dc19-4606-b744-34865bc6c5c5",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_hier = [sba_loans[['dset', 'LoanNr_ChkDgt', 'rand_cat']] \\\n",
    "              .set_index('LoanNr_ChkDgt').sort_index()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "cd6910ee-5ffe-4def-9aa8-7dba76446b08",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  107/21735 [..............................] - ETA: 30s"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-05-29 07:38:48.767069: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:114] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "21735/21735 [==============================] - 30s 1ms/step\n"
     ]
    }
   ],
   "source": [
    "all_predictions = this_model.predict([X.drop(columns='dset')] +\n",
    "                                     [Xn.drop(columns='dset') for Xn in X_hier])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "98a2cce1-b7d3-4ebc-8869-80705297e145",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_predictions_df = pd.DataFrame(all_predictions, index=X.index) \\\n",
    "    .set_axis(['predict_prob'], axis=1) \\\n",
    "    .reset_index() \\\n",
    "    .merge(sba_loans[['target', 'LoanNr_ChkDgt', 'dset', 'dset_naics_holdout', 'NAICS']], \n",
    "           on='LoanNr_ChkDgt')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "3fc7da9c-6bd2-4d1b-b77f-b62679825d5d",
   "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>predict_prob</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>predict_prob</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.283723</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>target</th>\n",
       "      <td>0.283723</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              predict_prob    target\n",
       "predict_prob      1.000000  0.283723\n",
       "target            0.283723  1.000000"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_predictions_df[['predict_prob', 'target']].corr(method='spearman')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e79afc4e-cb67-4064-9b19-dc52500baef2",
   "metadata": {},
   "source": [
    "##### Threshold Tune & Binary Predictions\n",
    "Using training probability predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "ccd0bc23-fbc0-40d0-8d8b-f2302405a875",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_pred_train = all_predictions_df[all_predictions_df['dset'] == 'train']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "df36d9e9-f35e-46b3-a5be-08081f998319",
   "metadata": {},
   "outputs": [],
   "source": [
    "thresh_tune_data = sbnn_metrics.get_f1_frame(all_pred_train['target'], \n",
    "                                        all_pred_train['predict_prob'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "de634156-caf6-4a52-b5d7-44b8e986aa1d",
   "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>thresh</th>\n",
       "      <th>f1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.22</td>\n",
       "      <td>0.430445</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.20</td>\n",
       "      <td>0.429469</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.24</td>\n",
       "      <td>0.428558</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    thresh        f1\n",
       "11    0.22  0.430445\n",
       "10    0.20  0.429469\n",
       "12    0.24  0.428558"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "thresh_tune_data.sort_values('f1', ascending=False, inplace=True)\n",
    "thresh_tune_data.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "c746b880-d781-4bc1-9df2-89211127c22a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.22"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_thresh = thresh_tune_data['thresh'].iloc[0]\n",
    "best_thresh"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3c0ac5e2-be0d-42f0-960d-1f81b4a5d4e2",
   "metadata": {},
   "source": [
    "##### Append binary predictions to probability predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "641da72e-1bca-4fe7-96c4-6c029b521bd7",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_predictions_df['predict_bin'] = sbnn_metrics.get_binary_predictions(all_predictions_df['predict_prob'], best_thresh)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "9f6ecdf4-ac01-4c39-8315-e8583daa337e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "predict_bin\n",
       "0    0.531229\n",
       "1    0.468771\n",
       "Name: proportion, dtype: float64"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_predictions_df['predict_bin'].value_counts(normalize=True, dropna=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "cbfed830-7ff8-40f6-a691-6b894b26d6bc",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_predictions_df.to_parquet(Path(setup.temp_path).joinpath('44_DATA_predictions.parquet'))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb0ecbd2-abff-46d8-8c70-f7bccd363b69",
   "metadata": {},
   "source": [
    "## Metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "be4ceefe-c7a8-45bd-8573-2daefe99a1c2",
   "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</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",
       "      <th>dset_naics_holdout</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>test</td>\n",
       "      <td>0.622243</td>\n",
       "      <td>0.427818</td>\n",
       "      <td>0.308019</td>\n",
       "      <td>0.700117</td>\n",
       "      <td>0.335895</td>\n",
       "      <td>0.700816</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>train</td>\n",
       "      <td>0.616053</td>\n",
       "      <td>0.430445</td>\n",
       "      <td>0.307507</td>\n",
       "      <td>0.717157</td>\n",
       "      <td>0.353049</td>\n",
       "      <td>0.706834</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>val</td>\n",
       "      <td>0.612652</td>\n",
       "      <td>0.426021</td>\n",
       "      <td>0.305190</td>\n",
       "      <td>0.705239</td>\n",
       "      <td>0.332132</td>\n",
       "      <td>0.695031</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>test</td>\n",
       "      <td>0.613267</td>\n",
       "      <td>0.424135</td>\n",
       "      <td>0.304057</td>\n",
       "      <td>0.700953</td>\n",
       "      <td>0.332994</td>\n",
       "      <td>0.692226</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>test</td>\n",
       "      <td>0.637161</td>\n",
       "      <td>0.434228</td>\n",
       "      <td>0.314996</td>\n",
       "      <td>0.698699</td>\n",
       "      <td>0.341772</td>\n",
       "      <td>0.714950</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    dset  accuracy_score  f1_score  precision_score  recall_score  \\\n",
       "0   test        0.622243  0.427818         0.308019      0.700117   \n",
       "1  train        0.616053  0.430445         0.307507      0.717157   \n",
       "2    val        0.612652  0.426021         0.305190      0.705239   \n",
       "0   test        0.613267  0.424135         0.304057      0.700953   \n",
       "1   test        0.637161  0.434228         0.314996      0.698699   \n",
       "\n",
       "   average_precision_score  roc_auc_score  dset_naics_holdout  \n",
       "0                 0.335895       0.700816                 NaN  \n",
       "1                 0.353049       0.706834                 NaN  \n",
       "2                 0.332132       0.695031                 NaN  \n",
       "0                 0.332994       0.692226                 0.0  \n",
       "1                 0.341772       0.714950                 1.0  "
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metrics_dset_df = all_predictions_df.groupby('dset') \\\n",
    "    .apply(lambda x: sbnn_metrics.dset_metrics(x.target, x.predict_bin, x.predict_prob)) \\\n",
    "    .reset_index()\n",
    "metrics_test_df = all_predictions_df[all_predictions_df['dset'] == 'test'] \\\n",
    "    .groupby(['dset', 'dset_naics_holdout']) \\\n",
    "    .apply(lambda x: sbnn_metrics.dset_metrics(x.target, x.predict_bin, x.predict_prob)) \\\n",
    "    .reset_index()\n",
    "metrics_df = pd.concat([metrics_dset_df, metrics_test_df])\n",
    "metrics_df.to_csv(Path(setup.temp_path).joinpath('44_REPORT_metrics.csv'), index=True)\n",
    "metrics_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "52e382b5-e9dc-47a1-9bf4-6d0b6b4a0c61",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6f69cf93-02e2-413b-8f2b-fcdb908057d3",
   "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
}