05_embed_rand.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "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": [
    "## Neural Network With NAICS Embedding Layer + Random Unseen Codes\n",
    "This is like 03, but the training data is randomized to inject the unseen embedding code (1) into the data."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e8d6ba37-5510-46a9-8de1-e87d3dfb6712",
   "metadata": {},
   "source": [
    "Add NAICS embedding to model.  Only add the baseline NAICS (not higher levels in the hierarchy).  "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e169c2e8-ba0f-488f-acf6-87d419bd4e99",
   "metadata": {},
   "source": [
    "Use the dataset with no missing values. "
   ]
  },
  {
   "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": 2,
   "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": 3,
   "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": 4,
   "id": "ae830c76-ee3d-4c70-870e-aec00a1fc79e",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "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 IPython.display import display, HTML\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "9f1e4f2e-0e45-4e28-8e9c-212e0bb51199",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sba_nn.sba_nn import sbnn_metrics, sbnn_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "da3d6e2c-7884-41dc-8e8d-6ea5747c5e5a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<module 'sba_nn.sba_nn.sbnn_model' from '/Users/valeriecarey/Documents/projects/2024_05_blog_nn_rand/code/sba_nn/sba_nn/sbnn_model.py'>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "importlib.reload(sbnn_model)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3cbc00d2-bc83-42d9-83eb-65fcf306dccd",
   "metadata": {},
   "source": [
    "## Input Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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": 9,
   "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": "code",
   "execution_count": 10,
   "id": "6f0ce28e-a80c-415b-b068-5c4d49370da8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'NAICS': 1170,\n",
       " 'NAICS_5': 764,\n",
       " 'NAICS_4': 345,\n",
       " 'NAICS_3': 107,\n",
       " 'NAICS_sector': 21}"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with open(Path(setup.temp_path).joinpath('01_DATA_naics_max_encodings.pkl'), 'rb') as fin:\n",
    "    naics_max_levels= pickle.load(fin)\n",
    "naics_max_levels"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4d4116da-d689-478a-9c5d-a330f231e615",
   "metadata": {},
   "source": [
    "##### Features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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": 12,
   "id": "a74fa386-4011-45ee-bdb5-d671db9ffa91",
   "metadata": {},
   "outputs": [],
   "source": [
    "# NAICS feature - just base\n",
    "features_naics = ['NAICS']\n",
    "features_naics_max_levels  = [naics_max_levels[n] for n in features_naics]\n",
    "features_naics_emb_dim = [setup.nn_naics_embed_size_dict[n] for n in features_naics]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "867031ee-8b6e-424d-b497-8e19409ba1d9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1170]\n",
      "[8]\n"
     ]
    }
   ],
   "source": [
    "print(features_naics_max_levels)\n",
    "print(features_naics_emb_dim)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3da44371-2bb8-4b1f-ba5f-dd08571365d9",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Data Adjustment\n",
    "Inject \"1\" values into train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "9813c1ba-0668-40f0-82fa-f04d3b514e3a",
   "metadata": {},
   "outputs": [],
   "source": [
    "for i, n in enumerate(features_naics):\n",
    "    adj_index, _ = train_test_split(sba_loans[sba_loans['dset'] == 'train'].index, \n",
    "                                    train_size=0.10, random_state=645 + i)\n",
    "    sba_loans.loc[adj_index, n] = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "61e7b587-cdf0-4074-810b-8462bb12d822",
   "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>NAICS</th>\n",
       "      <th>False</th>\n",
       "      <th>True</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dset</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>test</th>\n",
       "      <td>95552</td>\n",
       "      <td>57496</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>train</th>\n",
       "      <td>402164</td>\n",
       "      <td>44684</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>val</th>\n",
       "      <td>95593</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "NAICS   False  True \n",
       "dset                \n",
       "test    95552  57496\n",
       "train  402164  44684\n",
       "val     95593     11"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Show unseen counts - expect some in train now!\n",
    "pd.crosstab(sba_loans['dset'], sba_loans['NAICS'] == 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b979396c-c822-4940-a292-6e07f1b54f58",
   "metadata": {},
   "source": [
    "##### Datasets for train, validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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": 17,
   "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": 18,
   "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": 19,
   "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": 20,
   "id": "1b8ce4c9-6907-4b70-982a-b92a52b3159d",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_naics = [sba_loans[['dset', 'LoanNr_ChkDgt', n]].set_index('LoanNr_ChkDgt').sort_index() \\\n",
    "           for n in features_naics]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "a6f129e0-a4d0-4d93-9886-153c9803d403",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_naics_train = [Xn[Xn['dset'] == 'train'].drop(columns='dset') for Xn in X_naics]\n",
    "X_naics_val =  [Xn[Xn['dset'] == 'val'].drop(columns='dset') for Xn in X_naics]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a5c68623-4599-4472-a13a-9400d101633e",
   "metadata": {},
   "source": [
    "## Create, Fit Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "4d48ddfa-580c-4321-9917-0bd176b67d87",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-05-19 13:26:39.105160: I metal_plugin/src/device/metal_device.cc:1154] Metal device set to: Apple M1\n",
      "2024-05-19 13:26:39.105186: I metal_plugin/src/device/metal_device.cc:296] systemMemory: 16.00 GB\n",
      "2024-05-19 13:26:39.105195: I metal_plugin/src/device/metal_device.cc:313] maxCacheSize: 5.33 GB\n",
      "2024-05-19 13:26:39.105238: 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",
      "2024-05-19 13:26:39.105257: 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",
    "                                         features_naics_max_levels,\n",
    "                                         features_naics_emb_dim,\n",
    "                                         naics_embedding_names = features_naics,\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": 23,
   "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)                 9360      ['input_naics_00[0][0]']      \n",
      "                                                                                                  \n",
      " NAICS (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         ['NAICS[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: 20113 (78.57 KB)\n",
      "Trainable params: 20113 (78.57 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "this_model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "41bff593-4d39-44a3-861a-ef30cabbdb7f",
   "metadata": {},
   "outputs": [],
   "source": [
    "callback = keras.callbacks.EarlyStopping(patience=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "31772016-3643-468b-8d65-c456168cad6e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-05-19 13:26:40.006800: 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": [
      "13961/13964 [============================>.] - ETA: 0s - loss: 0.4641 - auc: 0.3320 - auc_1: 0.6959"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-05-19 13:29:55.138997: 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 [==============================] - 221s 16ms/step - loss: 0.4641 - auc: 0.3320 - auc_1: 0.6959 - val_loss: 0.4540 - val_auc: 0.3694 - val_auc_1: 0.7221\n",
      "Epoch 2/20\n",
      "13964/13964 [==============================] - 223s 16ms/step - loss: 0.4560 - auc: 0.3548 - auc_1: 0.7130 - val_loss: 0.4522 - val_auc: 0.3772 - val_auc_1: 0.7274\n",
      "Epoch 3/20\n",
      "13964/13964 [==============================] - 221s 16ms/step - loss: 0.4538 - auc: 0.3619 - auc_1: 0.7180 - val_loss: 0.4508 - val_auc: 0.3803 - val_auc_1: 0.7286\n",
      "Epoch 4/20\n",
      "13964/13964 [==============================] - 223s 16ms/step - loss: 0.4526 - auc: 0.3651 - auc_1: 0.7204 - val_loss: 0.4520 - val_auc: 0.3811 - val_auc_1: 0.7280\n",
      "Epoch 5/20\n",
      "13964/13964 [==============================] - 225s 16ms/step - loss: 0.4516 - auc: 0.3689 - auc_1: 0.7224 - val_loss: 0.4499 - val_auc: 0.3846 - val_auc_1: 0.7304\n",
      "Epoch 6/20\n",
      "13964/13964 [==============================] - 226s 16ms/step - loss: 0.4512 - auc: 0.3704 - auc_1: 0.7235 - val_loss: 0.4505 - val_auc: 0.3838 - val_auc_1: 0.7302\n",
      "Epoch 7/20\n",
      "13964/13964 [==============================] - 221s 16ms/step - loss: 0.4507 - auc: 0.3718 - auc_1: 0.7243 - val_loss: 0.4501 - val_auc: 0.3841 - val_auc_1: 0.7302\n",
      "Epoch 8/20\n",
      "13964/13964 [==============================] - 228s 16ms/step - loss: 0.4502 - auc: 0.3738 - auc_1: 0.7253 - val_loss: 0.4497 - val_auc: 0.3851 - val_auc_1: 0.7311\n",
      "Epoch 9/20\n",
      "13964/13964 [==============================] - 236s 17ms/step - loss: 0.4498 - auc: 0.3756 - auc_1: 0.7261 - val_loss: 0.4498 - val_auc: 0.3831 - val_auc_1: 0.7306\n",
      "Epoch 10/20\n",
      "13964/13964 [==============================] - 192s 14ms/step - loss: 0.4495 - auc: 0.3763 - auc_1: 0.7265 - val_loss: 0.4500 - val_auc: 0.3840 - val_auc_1: 0.7303\n",
      "Epoch 11/20\n",
      "13964/13964 [==============================] - 165s 12ms/step - loss: 0.4494 - auc: 0.3768 - auc_1: 0.7269 - val_loss: 0.4499 - val_auc: 0.3838 - val_auc_1: 0.7305\n"
     ]
    }
   ],
   "source": [
    "this_history = this_model.fit([X_train] + X_naics_train, y_train,\n",
    "                              validation_data=([X_val] + X_naics_val, y_val),\n",
    "                              batch_size=setup.nn_batch_size,\n",
    "                              epochs=setup.nn_epochs,\n",
    "                              callbacks=[callback])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "bd448bf9-89d7-489a-9575-de9e2138e678",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<module 'sba_nn.sba_nn.sbnn_model' from '/Users/valeriecarey/Documents/projects/2024_05_blog_nn_rand/code/sba_nn/sba_nn/sbnn_model.py'>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "importlib.reload(sbnn_model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "460c4b43-c673-4ee1-b070-d2d596ac4561",
   "metadata": {},
   "outputs": [],
   "source": [
    "this_history_df = sbnn_model.process_history(this_history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "ad4783f2-8d0b-4850-a45a-3a4be8e63318",
   "metadata": {},
   "outputs": [],
   "source": [
    "this_history_df.to_csv(Path(setup.temp_path).joinpath('05_REPORT_fit_history.csv'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "73e89efd-75a9-4b27-81a8-13e12d5482be",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 29,
     "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": 30,
   "id": "2342a31f-4891-4834-851a-b90b8b729c3c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 30,
     "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": 31,
   "id": "cee80637-cabf-4974-b476-08c5fba165ff",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 31,
     "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": 32,
   "id": "e184b0ca-1d31-4798-88a0-4e590b67062e",
   "metadata": {},
   "outputs": [],
   "source": [
    "this_model.save(Path(setup.temp_path).joinpath('05_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": 33,
   "id": "d5c39a5a-dc19-4606-b744-34865bc6c5c5",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_naics = [sba_loans[['dset', 'LoanNr_ChkDgt', n]].set_index('LoanNr_ChkDgt').sort_index() \\\n",
    "           for n in features_naics]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "cd6910ee-5ffe-4def-9aa8-7dba76446b08",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   93/21735 [..............................] - ETA: 35s"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-05-19 14:06:21.855697: 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 [==============================] - 37s 2ms/step\n"
     ]
    }
   ],
   "source": [
    "all_predictions = this_model.predict([X.drop(columns='dset')] +\n",
    "                                     [Xn.drop(columns='dset') for Xn in X_naics])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "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": 36,
   "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>0.321102</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>target</th>\n",
       "      <td>0.321102</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              predict_prob    target\n",
       "predict_prob      1.000000  0.321102\n",
       "target            0.321102  1.000000"
      ]
     },
     "execution_count": 36,
     "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": 37,
   "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": 38,
   "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": 39,
   "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>12</th>\n",
       "      <td>0.24</td>\n",
       "      <td>0.450816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.22</td>\n",
       "      <td>0.450790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.20</td>\n",
       "      <td>0.448934</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    thresh        f1\n",
       "12    0.24  0.450816\n",
       "11    0.22  0.450790\n",
       "10    0.20  0.448934"
      ]
     },
     "execution_count": 39,
     "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": 40,
   "id": "c746b880-d781-4bc1-9df2-89211127c22a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.24"
      ]
     },
     "execution_count": 40,
     "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": 41,
   "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": 42,
   "id": "9f6ecdf4-ac01-4c39-8315-e8583daa337e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "predict_bin\n",
       "0    0.596165\n",
       "1    0.403835\n",
       "Name: proportion, dtype: float64"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_predictions_df['predict_bin'].value_counts(normalize=True, dropna=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "cbfed830-7ff8-40f6-a691-6b894b26d6bc",
   "metadata": {},
   "outputs": [],
   "source": [
    "all_predictions_df.to_parquet(Path(setup.temp_path).joinpath('05_DATA_predictions.parquet'))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb0ecbd2-abff-46d8-8c70-f7bccd363b69",
   "metadata": {},
   "source": [
    "## Metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "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>0.663308</td>\n",
       "      <td>0.441143</td>\n",
       "      <td>0.331594</td>\n",
       "      <td>0.658785</td>\n",
       "      <td>0.369665</td>\n",
       "      <td>0.723398</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>train</td>\n",
       "      <td>0.666356</td>\n",
       "      <td>0.450816</td>\n",
       "      <td>0.337943</td>\n",
       "      <td>0.676903</td>\n",
       "      <td>0.388130</td>\n",
       "      <td>0.733230</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>val</td>\n",
       "      <td>0.667252</td>\n",
       "      <td>0.451025</td>\n",
       "      <td>0.339773</td>\n",
       "      <td>0.670601</td>\n",
       "      <td>0.383961</td>\n",
       "      <td>0.730563</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>test</td>\n",
       "      <td>0.665686</td>\n",
       "      <td>0.445767</td>\n",
       "      <td>0.336089</td>\n",
       "      <td>0.661705</td>\n",
       "      <td>0.378818</td>\n",
       "      <td>0.726329</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>test</td>\n",
       "      <td>0.659355</td>\n",
       "      <td>0.433432</td>\n",
       "      <td>0.324159</td>\n",
       "      <td>0.653836</td>\n",
       "      <td>0.352953</td>\n",
       "      <td>0.718785</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        0.663308  0.441143         0.331594      0.658785   \n",
       "1  train        0.666356  0.450816         0.337943      0.676903   \n",
       "2    val        0.667252  0.451025         0.339773      0.670601   \n",
       "0   test        0.665686  0.445767         0.336089      0.661705   \n",
       "1   test        0.659355  0.433432         0.324159      0.653836   \n",
       "\n",
       "   average_precision_score  roc_auc_score  dset_naics_holdout  \n",
       "0                 0.369665       0.723398                 NaN  \n",
       "1                 0.388130       0.733230                 NaN  \n",
       "2                 0.383961       0.730563                 NaN  \n",
       "0                 0.378818       0.726329                 0.0  \n",
       "1                 0.352953       0.718785                 1.0  "
      ]
     },
     "execution_count": 44,
     "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('05_REPORT_metrics.csv'), index=True)\n",
    "metrics_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "52e382b5-e9dc-47a1-9bf4-6d0b6b4a0c61",
   "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
}