Fix SARIMA and LSTM deployment issues

app.py

@@ -9,46 +9,41 @@ from sklearn.metrics import mean_absolute_error, mean_squared_error
 # Load the dataset
 webtraffic_data = pd.read_csv("webtraffic.csv")
 
-# Rename 'Hour Index' for easier use
-webtraffic_data.rename(columns={"Hour Index": "Datetime"}, inplace=True)
+# Convert 'Hour Index' to datetime
+start_date = pd.Timestamp("2024-01-01 00:00:00")
+webtraffic_data['Datetime'] = start_date + pd.to_timedelta(webtraffic_data['Hour Index'], unit='h')
+webtraffic_data.drop(columns=['Hour Index'], inplace=True)
 
-# Create a datetime-like index for visualization purposes
-webtraffic_data['Datetime'] = pd.date_range(start='2023-01-01', periods=len(webtraffic_data), freq='H')
-
-# Split the data into train/test for evaluation
+# Split the data into train/test
 train_size = int(len(webtraffic_data) * 0.8)
-test_size = len(webtraffic_data) - train_size
 train_data = webtraffic_data.iloc[:train_size]
 test_data = webtraffic_data.iloc[train_size:]
 
-# Load the pre-trained models
+# Load pre-trained models
 sarima_model = joblib.load("sarima_model.pkl")  # SARIMA model
 lstm_model = tf.keras.models.load_model("lstm_model.keras")  # LSTM model
 
-# Initialize future periods for prediction
-future_periods = len(test_data)
-
-# Generate predictions for SARIMA
-sarima_predictions = sarima_model.forecast(steps=future_periods)
-
-# Prepare data for LSTM predictions
+# Initialize scalers and scale the data for LSTM
 from sklearn.preprocessing import MinMaxScaler
 
 scaler_X = MinMaxScaler(feature_range=(0, 1))
 scaler_y = MinMaxScaler(feature_range=(0, 1))
 
-# Fit the scaler to the training data
+# Fit scalers on the training data
 X_train_scaled = scaler_X.fit_transform(train_data['Sessions'].values.reshape(-1, 1))
 y_train_scaled = scaler_y.fit_transform(train_data['Sessions'].values.reshape(-1, 1))
 
-# Scale test data
+# Scale the test data
 X_test_scaled = scaler_X.transform(test_data['Sessions'].values.reshape(-1, 1))
 y_test_scaled = scaler_y.transform(test_data['Sessions'].values.reshape(-1, 1))
 
-# Reshape data for LSTM input
+# Reshape test data for LSTM
 X_test_lstm = X_test_scaled.reshape((X_test_scaled.shape[0], 1, X_test_scaled.shape[1]))
 
-# Predict with LSTM
+# Generate predictions for SARIMA
+sarima_predictions = sarima_model.predict(start=len(train_data), end=len(webtraffic_data) - 1)
+
+# Generate predictions for LSTM
 lstm_predictions_scaled = lstm_model.predict(X_test_lstm)
 lstm_predictions = scaler_y.inverse_transform(lstm_predictions_scaled).flatten()
 
@@ -59,19 +54,18 @@ future_predictions = pd.DataFrame({
     "LSTM_Predicted": lstm_predictions
 })
 
-# Calculate metrics for both models
-mae_sarima_future = mean_absolute_error(test_data['Sessions'], sarima_predictions)
-rmse_sarima_future = mean_squared_error(test_data['Sessions'], sarima_predictions, squared=False)
+# Calculate metrics
+mae_sarima = mean_absolute_error(test_data['Sessions'], sarima_predictions)
+rmse_sarima = mean_squared_error(test_data['Sessions'], sarima_predictions, squared=False)
 
-mae_lstm_future = mean_absolute_error(test_data['Sessions'], lstm_predictions)
-rmse_lstm_future = mean_squared_error(test_data['Sessions'], lstm_predictions, squared=False)
+mae_lstm = mean_absolute_error(test_data['Sessions'], lstm_predictions)
+rmse_lstm = mean_squared_error(test_data['Sessions'], lstm_predictions, squared=False)
 
-# Function to generate plot based on the selected model
+# Function to generate plots
 def generate_plot(model):
     """Generate plot based on the selected model."""
     plt.figure(figsize=(15, 6))
-    actual_dates = test_data['Datetime']
-    plt.plot(actual_dates, test_data['Sessions'], label='Actual Traffic', color='black', linestyle='dotted', linewidth=2)
+    plt.plot(test_data['Datetime'], test_data['Sessions'], label='Actual Traffic', color='black', linestyle='dotted', linewidth=2)
 
     if model == "SARIMA":
         plt.plot(future_predictions['Datetime'], future_predictions['SARIMA_Predicted'], label='SARIMA Predicted', color='blue', linewidth=2)
@@ -89,41 +83,41 @@ def generate_plot(model):
     plt.close()
     return plot_path
 
-# Function to display metrics for both models
+# Function to display metrics
 def display_metrics():
-    """Generate a DataFrame with metrics for SARIMA and LSTM."""
+    """Generate metrics for both models."""
     metrics = {
         "Model": ["SARIMA", "LSTM"],
-        "Mean Absolute Error (MAE)": [mae_sarima_future, mae_lstm_future],
-        "Root Mean Squared Error (RMSE)": [rmse_sarima_future, rmse_lstm_future]
+        "Mean Absolute Error (MAE)": [mae_sarima, mae_lstm],
+        "Root Mean Squared Error (RMSE)": [rmse_sarima, rmse_lstm]
     }
     return pd.DataFrame(metrics)
 
 # Gradio interface function
 def dashboard_interface(model="SARIMA"):
     """Generate plot and metrics for the selected model."""
-    plot_path = generate_plot(model)  # Generate plot for the selected model
-    metrics_df = display_metrics()   # Get metrics
+    plot_path = generate_plot(model)
+    metrics_df = display_metrics()
     return plot_path, metrics_df.to_string()
 
-# Build the Gradio interface
+# Build the Gradio dashboard
 with gr.Blocks() as dashboard:
-    gr.Markdown("## Interactive Web Traffic Prediction Dashboard")
-    gr.Markdown("Use the dropdown menu to select a model and view its predictions vs actual traffic along with performance metrics.")
+    gr.Markdown("## Web Traffic Prediction Dashboard")
+    gr.Markdown("Select a model to view its predictions and performance metrics.")
 
     # Dropdown for model selection
     model_selection = gr.Dropdown(["SARIMA", "LSTM"], label="Select Model", value="SARIMA")
 
     # Outputs: Plot and Metrics
     plot_output = gr.Image(label="Prediction Plot")
-    metrics_output = gr.Textbox(label="Metrics", lines=15)
+    metrics_output = gr.Textbox(label="Metrics", lines=10)
 
     # Button to update dashboard
     gr.Button("Update Dashboard").click(
-        fn=dashboard_interface,        # Function to call
-        inputs=[model_selection],      # Inputs to the function
-        outputs=[plot_output, metrics_output]  # Outputs from the function
+        fn=dashboard_interface,
+        inputs=[model_selection],
+        outputs=[plot_output, metrics_output]
     )
 
-# Launch the Gradio dashboard
+# Launch the dashboard
 dashboard.launch()