Update app.py

app.py

@@ -1,50 +1,45 @@
-import gradio as gr
 import joblib
 import pandas as pd
+import gradio as gr
 
-# Load model artifacts
-model = joblib.load('fall_detection_model.joblib')
+# Load model, scaler, feature names etc.
+model = joblib.load('model.joblib')
 scaler = joblib.load('scaler.joblib')
-encoder = joblib.load('encoder.joblib')
-feature_names = joblib.load('feature_names.joblib')  # your full feature list
+feature_names = joblib.load('feature_names.joblib')  # list of all features in correct order
 
-# Extract categories for dropdown menus
-movement_choices = encoder.categories_[0].tolist()
-location_choices = encoder.categories_[1].tolist()
-day_choices = [col.replace('day_of_week_', '') for col in feature_names if col.startswith('day_of_week_')]
+# For dropdown options, extract from encoder info or hardcode:
+movement_activities = ['Lying', 'No Movement', 'Sitting', 'Walking']
+locations = ['Bathroom', 'Bedroom', 'Kitchen', 'Living Room']
+days_of_week = ['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday']
 
 def predict_fall(movement_activity, location, day_of_week, hour_of_day, minute_of_day, time_since_last_event):
     try:
-        # 1. Initialize all features to zero using model's expected order
+        # Initialize zero data dict for all features
         data = {f: 0 for f in feature_names}
 
-        # 2. Set one-hot encoded categorical features
+        # One-hot encode categorical features
         data[f'Movement Activity_{movement_activity}'] = 1
         data[f'Location_{location}'] = 1
         data[f'day_of_week_{day_of_week}'] = 1
 
-        # 3. Set numeric features
+        # Set numeric features
         data['hour_of_day'] = hour_of_day
         data['minute_of_day'] = minute_of_day
         data['time_since_last_event'] = time_since_last_event
 
-        # 4. Create DataFrame with model feature order
-        input_df = pd.DataFrame([data], columns=feature_names)
+        # Create DataFrame with float dtype to avoid warnings
+        input_df = pd.DataFrame([data], columns=feature_names, dtype=float)
 
-        # 5. Select scaler columns in scaler's order
+        # Scale numeric features only (assumes scaler was fit on these)
         scaler_cols = scaler.feature_names_in_
-        scaler_df = input_df[scaler_cols]
+        scaled_features = scaler.transform(input_df[scaler_cols])
+        input_df.loc[:, scaler_cols] = scaled_features
 
-        # 6. Transform numeric/scaler features
-        scaled_features = scaler.transform(scaler_df)
+        # Ensure columns are in model's expected order
+        input_df = input_df[model.feature_names_in_]
 
-        # 7. Replace scaler columns in input_df (which is in model order) with scaled data
-        # Use .loc to avoid dtype warning, cast scaled_features to float explicitly
-        input_df.loc[:, scaler_cols] = scaled_features.astype(float)
-
-        # 8. Now predict using the model, passing full input_df in model feature order
+        # Predict probability
         pred_proba = model.predict_proba(input_df)[0, 1]
-
         threshold = 0.4
         pred_label = "Fall Detected" if pred_proba >= threshold else "No Fall"
 
@@ -55,22 +50,29 @@ def predict_fall(movement_activity, location, day_of_week, hour_of_day, minute_o
         traceback.print_exc()
         return f"Error: {str(e)}. Check server logs."
 
+# Build Gradio interface
+with gr.Blocks() as demo:
+    gr.Markdown("## Fall Prediction")
+
+    with gr.Row():
+        movement_input = gr.Dropdown(choices=movement_activities, label="Movement Activity")
+        location_input = gr.Dropdown(choices=locations, label="Location")
+        day_input = gr.Dropdown(choices=days_of_week, label="Day of Week")
+
+    with gr.Row():
+        hour_input = gr.Slider(minimum=0, maximum=23, step=1, label="Hour of Day")
+        minute_input = gr.Slider(minimum=0, maximum=59, step=1, label="Minute of Day")
+
+    time_since_input = gr.Number(label="Time Since Last Event (minutes)")
+
+    predict_button = gr.Button("Predict")
+    output = gr.Textbox(label="Prediction Result")
 
-# Gradio interface setup
-iface = gr.Interface(
-    fn=predict_fall,
-    inputs=[
-        gr.Dropdown(movement_choices, label="Movement Activity"),
-        gr.Dropdown(location_choices, label="Location"),
-        gr.Dropdown(day_choices, label="Day of Week"),
-        gr.Slider(0, 23, step=1, label="Hour of Day"),
-        gr.Slider(0, 59, step=1, label="Minute of Day"),
-        gr.Number(label="Time Since Last Event (seconds)", value=60),
-    ],
-    outputs="text",
-    title="Fall Detection Model",
-    description="Predicts if a fall is detected based on sensor input."
-)
+    predict_button.click(
+        predict_fall,
+        inputs=[movement_input, location_input, day_input, hour_input, minute_input, time_since_input],
+        outputs=output
+    )
 
 if __name__ == "__main__":
-    iface.launch()
+    demo.launch()