Update app.py

app.py

@@ -1,12 +1,11 @@
 import gradio as gr
 import pandas as pd
-import numpy as np
 
 from sklearn.datasets import (
     load_iris,
     load_breast_cancer,
     fetch_california_housing,
-    load_diabetes
+    load_diabetes,
 )
 
 from sklearn.model_selection import train_test_split
@@ -14,110 +13,137 @@ from sklearn.preprocessing import StandardScaler
 
 # Classification Models
 from sklearn.linear_model import LogisticRegression
-from sklearn.svm import SVC
 from sklearn.tree import DecisionTreeClassifier
 from sklearn.ensemble import RandomForestClassifier
+from sklearn.neighbors import KNeighborsClassifier
 
 # Regression Models
 from sklearn.linear_model import LinearRegression
-from sklearn.svm import SVR
 from sklearn.tree import DecisionTreeRegressor
 from sklearn.ensemble import RandomForestRegressor
+from sklearn.svm import SVR
 
 # Metrics
 from sklearn.metrics import (
     accuracy_score,
     f1_score,
     mean_squared_error,
-    r2_score
+    r2_score,
 )
 
+# =====================================================
+# TITANIC DATASET
+# =====================================================
 
-# ==================================================
-# Load Dataset
-# ==================================================
-
-def load_dataset(task_type, dataset_name):
+def load_titanic():
+    url = "https://raw.githubusercontent.com/datasciencedojo/datasets/master/titanic.csv"
+    df = pd.read_csv(url)
 
-    # Classification datasets
-    if task_type == "classification":
+    df = df[["Pclass", "Sex", "Age", "Fare", "Survived"]]
 
-        if dataset_name == "Iris":
-            data = load_iris(as_frame=True)
-
-        elif dataset_name == "Breast Cancer":
-            data = load_breast_cancer(as_frame=True)
+    df["Age"] = df["Age"].fillna(df["Age"].mean())
+    df["Sex"] = df["Sex"].map({"male": 0, "female": 1})
 
-        else:
-            return None, None
+    X = df.drop("Survived", axis=1)
+    y = df["Survived"]
 
-    # Regression datasets
-    else:
+    return X, y
 
-        if dataset_name == "California Housing":
-            data = fetch_california_housing(as_frame=True)
 
-        elif dataset_name == "Diabetes":
-            data = load_diabetes(as_frame=True)
+# =====================================================
+# BOSTON DATASET
+# =====================================================
 
-        else:
-            return None, None
+def load_boston():
+    url = "https://raw.githubusercontent.com/selva86/datasets/master/BostonHousing.csv"
+    df = pd.read_csv(url)
 
-    X = data.data
-    y = data.target
+    X = df.drop("medv", axis=1)
+    y = df["medv"]
 
     return X, y
 
 
-# ==================================================
-# Main Function
-# ==================================================
+# =====================================================
+# MAIN FUNCTION
+# =====================================================
 
 def run_models(task_type, dataset_name):
 
-    # Load dataset
-    X, y = load_dataset(task_type, dataset_name)
+    # =========================
+    # CLASSIFICATION DATASETS
+    # =========================
 
-    if X is None:
-        return "Invalid dataset selection", ""
+    if task_type == "Classification":
 
-    # Split
-    X_train, X_test, y_train, y_test = train_test_split(
-        X,
-        y,
-        test_size=0.2,
-        random_state=42
-    )
-
-    # Scaling
-    scaler = StandardScaler()
+        if dataset_name == "Iris":
+            data = load_iris()
+            X = pd.DataFrame(data.data, columns=data.feature_names)
+            y = data.target
 
-    X_train = scaler.fit_transform(X_train)
-    X_test = scaler.transform(X_test)
+        elif dataset_name == "Breast Cancer":
+            data = load_breast_cancer()
+            X = pd.DataFrame(data.data, columns=data.feature_names)
+            y = data.target
 
-    # Classification Models
-    if task_type == "classification":
+        elif dataset_name == "Titanic":
+            X, y = load_titanic()
 
         models = {
             "Logistic Regression": LogisticRegression(max_iter=1000),
-            "SVM": SVC(),
             "Decision Tree": DecisionTreeClassifier(),
-            "Random Forest": RandomForestClassifier()
+            "Random Forest": RandomForestClassifier(),
+            "KNN": KNeighborsClassifier(),
         }
 
-    # Regression Models
+    # =========================
+    # REGRESSION DATASETS
+    # =========================
+
     else:
 
+        if dataset_name == "California Housing":
+            data = fetch_california_housing()
+            X = pd.DataFrame(data.data, columns=data.feature_names)
+            y = data.target
+
+        elif dataset_name == "Diabetes":
+            data = load_diabetes()
+            X = pd.DataFrame(data.data, columns=data.feature_names)
+            y = data.target
+
+        elif dataset_name == "Boston Housing":
+            X, y = load_boston()
+
         models = {
             "Linear Regression": LinearRegression(),
-            "SVR": SVR(),
             "Decision Tree": DecisionTreeRegressor(),
-            "Random Forest": RandomForestRegressor()
+            "Random Forest": RandomForestRegressor(),
+            "SVR": SVR(),
         }
 
+    # =========================
+    # SPLIT + SCALE
+    # =========================
+
+    X_train, X_test, y_train, y_test = train_test_split(
+        X,
+        y,
+        test_size=0.2,
+        random_state=42,
+    )
+
+    scaler = StandardScaler()
+
+    X_train = scaler.fit_transform(X_train)
+    X_test = scaler.transform(X_test)
+
+    # =========================
+    # TRAIN MODELS
+    # =========================
+
     results = []
 
-    # Train Models
     for name, model in models.items():
 
         model.fit(X_train, y_train)
@@ -125,37 +151,26 @@ def run_models(task_type, dataset_name):
         predictions = model.predict(X_test)
 
         # Classification Metrics
-        if task_type == "classification":
+        if task_type == "Classification":
 
             accuracy = accuracy_score(y_test, predictions)
+            f1 = f1_score(y_test, predictions, average="weighted")
 
-            f1 = f1_score(
-                y_test,
-                predictions,
-                average="weighted"
-            )
-
-            results.append([
-                name,
-                round(accuracy, 4),
-                round(f1, 4)
-            ])
+            results.append([name, accuracy, f1])
 
         # Regression Metrics
         else:
 
             mse = mean_squared_error(y_test, predictions)
-
             r2 = r2_score(y_test, predictions)
 
-            results.append([
-                name,
-                round(mse, 4),
-                round(r2, 4)
-            ])
+            results.append([name, mse, r2])
+
+    # =========================
+    # RESULTS TABLE
+    # =========================
 
-    # Results DataFrame
-    if task_type == "classification":
+    if task_type == "Classification":
 
         results_df = pd.DataFrame(
             results,
@@ -163,8 +178,9 @@ def run_models(task_type, dataset_name):
         )
 
         best_model = results_df.loc[
-            results_df["Accuracy"].idxmax()
-        ]["Model"]
+            results_df["Accuracy"].idxmax(),
+            "Model"
+        ]
 
     else:
 
@@ -174,24 +190,26 @@ def run_models(task_type, dataset_name):
         )
 
         best_model = results_df.loc[
-            results_df["R2 Score"].idxmax()
-        ]["Model"]
+            results_df["MSE"].idxmin(),
+            "Model"
+        ]
 
     return results_df, f"🏆 Best Model: {best_model}"
 
 
-# ==================================================
-# Update Dataset Choices
-# ==================================================
+# =====================================================
+# UPDATE DATASET OPTIONS
+# =====================================================
 
 def update_datasets(task_type):
 
-    if task_type == "classification":
+    if task_type == "Classification":
 
         return gr.Dropdown(
             choices=[
                 "Iris",
-                "Breast Cancer"
+                "Breast Cancer",
+                "Titanic"
             ],
             value="Iris"
         )
@@ -201,33 +219,35 @@ def update_datasets(task_type):
         return gr.Dropdown(
             choices=[
                 "California Housing",
-                "Diabetes"
+                "Diabetes",
+                "Boston Housing"
             ],
             value="California Housing"
         )
 
 
-# ==================================================
-# Gradio Interface
-# ==================================================
+# =====================================================
+# GRADIO UI
+# =====================================================
 
 with gr.Blocks() as demo:
 
-    gr.Markdown("# ML Model Comparison Tool")
+    gr.Markdown("# AI Model Comparison App")
 
     task_type = gr.Radio(
-        ["classification", "regression"],
-        label="Select Task Type",
-        value="classification"
+        choices=["Classification", "Regression"],
+        value="Classification",
+        label="Select Task Type"
     )
 
     dataset_name = gr.Dropdown(
         choices=[
             "Iris",
-            "Breast Cancer"
+            "Breast Cancer",
+            "Titanic"
         ],
-        label="Select Dataset",
-        value="Iris"
+        value="Iris",
+        label="Select Dataset"
     )
 
     task_type.change(
@@ -238,15 +258,14 @@ with gr.Blocks() as demo:
 
     run_button = gr.Button("Run Models")
 
-    results_output = gr.Dataframe(label="Results")
+    output_table = gr.Dataframe()
 
-    best_model_output = gr.Textbox(label="Best Model")
+    output_text = gr.Textbox()
 
     run_button.click(
         fn=run_models,
         inputs=[task_type, dataset_name],
-        outputs=[results_output, best_model_output]
+        outputs=[output_table, output_text]
     )
 
-
 demo.launch()