From d6e28b81f46799219a1ba2041c12d96c853a4851 Mon Sep 17 00:00:00 2001
From: Sverre Nystad <sverrenystad@gmail.com>
Date: Fri, 29 Mar 2024 21:20:10 +0100
Subject: [PATCH] feat: Add dependencies and update code for AutoGluon model

---
 models/auto_gluon.ipynb | 28 +++++++++++++++-------------
 1 file changed, 15 insertions(+), 13 deletions(-)

diff --git a/models/auto_gluon.ipynb b/models/auto_gluon.ipynb
index ee84bcb..9519020 100644
--- a/models/auto_gluon.ipynb
+++ b/models/auto_gluon.ipynb
@@ -43,6 +43,7 @@
    "outputs": [],
    "source": [
     "%pip install autogluon\n",
+    "%pip install bokeh\n",
     "%pip install scikit-learn"
    ]
   },
@@ -61,15 +62,15 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "%autoreload \n",
+    "%autoreload 2\n",
     "\n",
     "from sklearn.metrics import accuracy_score, classification_report\n",
     "import autogluon.core as ag\n",
+    "import pandas as pd\n",
     "from autogluon.tabular import TabularDataset, TabularPredictor\n",
     "\n",
-    "from src.features.post_processor import save_predictions\n",
-    "from src.ml_service import prepare_data, prepare_test_data\n",
-    "from src.config import TARGET_FEATURES"
+    "from src.ml_service import prepare_data, prepare_test_data, save_predictions\n",
+    "from src.config import TARGET_FEATURE"
    ]
   },
   {
@@ -86,10 +87,11 @@
    "outputs": [],
    "source": [
     "x_train, _, x_test, y_train, _, y_test = prepare_data(validation_size=0, test_size=0.1)\n",
-    "for target_feature_name in TARGET_FEATURES:\n",
-    "    x_train[target_feature_name] = y_train\n",
     "\n",
-    "data = TabularDataset(x_train)"
+    "combined_train_data = pd.concat([x_train, y_train], axis=1)\n",
+    "combined_test_data = pd.concat([x_test, y_test], axis=1)\n",
+    "training_data = TabularDataset(combined_train_data)\n",
+    "test_data = TabularDataset(combined_test_data)"
    ]
   },
   {
@@ -122,9 +124,9 @@
    "outputs": [],
    "source": [
     "# Initialize the AutoGluon TabularPredictor\n",
-    "time_limit = 24*60*60 # Set this to longest time you are willing to wait (in seconds)\n",
-    "metric = 'roc_auc'\n",
-    "predictor = TabularPredictor(label=target_feature_name, eval_metric=metric).fit(data, time_limit=time_limit, presets='best_quality')"
+    "time_limit = 3*60 #24*60*60 # Set this to longest time you are willing to wait (in seconds)\n",
+    "metric = 'log_loss'\n",
+    "predictor = TabularPredictor(label=TARGET_FEATURE, eval_metric=metric).fit(training_data, time_limit=time_limit, presets='best_quality')"
    ]
   },
   {
@@ -166,7 +168,7 @@
     "test_accuracy = accuracy_score(y_test, y_test_pred)\n",
     "print(\"Test Accuracy: \", test_accuracy)\n",
     "print(\"Test Classification Report:\\n\", classification_report(y_test, y_test_pred))\n",
-    "# predictor.leaderboard(x_test, silent=True)\n"
+    "predictor.leaderboard(test_data)\n"
    ]
   },
   {
@@ -182,8 +184,8 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "x_test = prepare_test_data()\n",
-    "final_predictions = predictor.predict(x_test)"
+    "x_final_test = prepare_test_data()\n",
+    "final_predictions = predictor.predict(x_final_test)"
    ]
   },
   {