Update pages/1_User_Defined_DataLab.py

pages/1_User_Defined_DataLab.py

@@ -10,17 +10,17 @@ from sklearn.model_selection import train_test_split
 from mlxtend.plotting import plot_decision_regions
 import numpy as np
 
+st.set_page_config(layout="wide")
 st.title("🧠 Neural Network Playground - Custom Dataset")
 
-# User input parameters
+# Generate synthetic 2-feature dataset
 n_samples = st.slider("Number of Samples", 100, 1000, 300)
 noise = st.slider("Noise Level", 0.0, 1.0, 0.2)
 random_state = st.number_input("Random State", value=42)
 
-# Generate synthetic 2-feature data
 X, y = make_classification(n_samples=n_samples, n_features=2, n_redundant=0,
-                           n_informative=2, n_clusters_per_class=1, flip_y=noise, 
-                           random_state=random_state, class_sep=4)
+                           n_informative=2, n_clusters_per_class=1,
+                           flip_y=noise, random_state=random_state, class_sep=4)
 
 df = pd.DataFrame(X, columns=["X1", "X2"])
 df["label"] = y
@@ -28,49 +28,70 @@ df["label"] = y
 st.write("### 📄 Preview of Generated Data")
 st.dataframe(df.head())
 
-st.session_state['X'] = X
-st.session_state['y'] = y
-
 # Scatter plot
 st.write("### 🎯 Feature Scatter Plot by Class Label")
 fig, ax = plt.subplots()
 sns.scatterplot(data=df, x="X1", y="X2", hue="label", palette="deep", ax=ax)
 st.pyplot(fig)
 
-# Sidebar - Network Config
-st.sidebar.title("🛠️ Network Configuration")
+# Sidebar - Model Configuration
+st.sidebar.title("🛠️ Neural Network Configuration")
 n_layers = st.sidebar.slider("Number of Hidden Layers", 1, 4, 2)
 hidden_activation = st.sidebar.selectbox("Activation for Hidden Layers", ["tanh", "sigmoid"])
 
 hidden_config = []
 for i in range(n_layers):
-    units = st.sidebar.slider(f"Neurons in Layer {i+1}", 1, 10, 4)
+    units = st.sidebar.slider(f"Neurons in Layer {i+1}", 1, 10, 4, key=f"layer_{i}")
     hidden_config.append(units)
 
 learning_rate = st.sidebar.slider("Learning Rate", 0.0001, 1.0, 0.1, step=0.0001, format="%.4f")
-batch_size = st.sidebar.slider("Batch Size", 1, 1000, 200)
+batch_size = st.sidebar.slider("Batch Size", 1, 512, 64)
 epochs = st.sidebar.slider("Epochs", 100, 2000, 350, step=50)
 
 # Train model
 if st.button("🚀 Train Model"):
-    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
+    with st.spinner("Training in progress... please wait ⏳"):
+        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
+
+        # Build model
+        model = Sequential()
+        model.add(InputLayer(shape=(2,)))
+        for units in hidden_config:
+            model.add(Dense(units, activation=hidden_activation))
+        model.add(Dense(1, activation="sigmoid"))
+
+        model.compile(optimizer=SGD(learning_rate=learning_rate),
+                      loss='binary_crossentropy',
+                      metrics=['accuracy'])
 
-    model = Sequential()
-    model.add(InputLayer(shape=(2,)))
-    for units in hidden_config:
-        model.add(Dense(units, activation=hidden_activation))
-    model.add(Dense(1, activation="sigmoid"))
+        history = model.fit(X_train, y_train, batch_size=batch_size,
+                            epochs=epochs, validation_split=0.2, verbose=0)
 
-    model.compile(optimizer=SGD(learning_rate=learning_rate),
-                  loss='binary_crossentropy',
-                  metrics=['accuracy'])
+    # ✅ Keras model wrapper for mlxtend
+    class KerasClassifierWrapper:
+        def __init__(self, model):
+            self.model = model
 
-    history = model.fit(X_train, y_train, batch_size=batch_size, epochs=epochs, validation_split=0.2, verbose=0)
+        def predict(self, X):
+            return (self.model.predict(X) > 0.5).astype(int).flatten()
 
-    # Plot decision region
-    st.write("### 🧭 Decision Boundary")
+    wrapped_model = KerasClassifierWrapper(model)
+
+    # 🎯 Plot decision region
+    st.subheader("🧭 Decision Region")
     fig2, ax2 = plt.subplots()
-    plot_decision_regions(X=X, y=y.astype(int), 
-                          clf=lambda X_: (model.predict(X_) > 0.5).astype(int).flatten(),
-                          legend=2, ax=ax2)
+    plot_decision_regions(X=X, y=y.astype(int),
+                          clf=wrapped_model, legend=2, ax=ax2)
     st.pyplot(fig2)
+
+    # 📉 Plot training history
+    st.subheader("📈 Training & Validation Loss")
+    fig3, ax3 = plt.subplots()
+    ax3.plot(history.history['loss'], label='Train Loss')
+    ax3.plot(history.history['val_loss'], label='Val Loss')
+    ax3.set_xlabel("Epochs")
+    ax3.set_ylabel("Loss")
+    ax3.legend()
+    st.pyplot(fig3)
+
+    st.success("✅ Model training completed!")