Update app.py

app.py

@@ -7,8 +7,8 @@ from app.ml_logic import get_combined_prediction
 # --- Description for the Gradio Interface ---
 title = "Parkinson's Disease Detection System"
 description = """
-Upload a voice recording and input symptom information to get a prediction for Parkinson's disease. 
-This tool uses a combination of a Convolutional Neural Network (CNN) on a voice spectrogram 
+Upload a voice recording and input symptom information to get a prediction for Parkinson's disease.
+This tool uses a combination of a Convolutional Neural Network (CNN) on a voice spectrogram
 and a machine learning model on user-reported symptoms.
 """
 
@@ -18,20 +18,16 @@ def predict(tremor, stiffness, walking_issue, age, audio_input):
     This function takes inputs from the Gradio interface, processes them,
     and calls the core machine learning logic.
     """
-    # 1. Handle the symptom data (convert boolean from checkboxes to 0/1)
+    # 1. Handle the symptom data (convert "Yes"/"No" from radio buttons to 0/1)
+    # THIS SECTION IS MODIFIED
     symptom_data = {
-        'tremor': 1 if tremor else 0,
-        'stiffness': 1 if stiffness else 0,
-        'walking_issue': 1 if walking_issue else 0
+        'tremor': 1 if tremor == "Yes" else 0,
+        'stiffness': 1 if stiffness == "Yes" else 0,
+        'walking_issue': 1 if walking_issue == "Yes" else 0
     }
 
     # 2. Handle the audio data
-    # Gradio provides the audio as a tuple (sample_rate, numpy_array)
-    # Our ml_logic function expects a file path. So, we save it to a temporary file.
     sample_rate, audio_data = audio_input
-    
-    # Create a temporary file to store the audio
-    # Using a 'with' statement ensures the file is deleted automatically afterward
     with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as temp_audio_file:
         sf.write(temp_audio_file.name, audio_data, sample_rate)
         temp_audio_path = temp_audio_file.name
@@ -43,7 +39,6 @@ def predict(tremor, stiffness, walking_issue, age, audio_input):
     # 3. Call your existing ML logic function
     try:
         final_label, raw_label, score = get_combined_prediction(symptom_data, temp_audio_path, age)
-        # Format the score to be more readable
         score_percentage = f"{score:.2%}"
         return final_label, raw_label, score_percentage
     except Exception as e:
@@ -57,16 +52,18 @@ def predict(tremor, stiffness, walking_issue, age, audio_input):
 
 
 # --- Define the Gradio Interface Components ---
+# THIS SECTION IS MODIFIED
 inputs = [
-    gr.Checkbox(label="Do you experience tremors at rest?"),
-    gr.Checkbox(label="Do you experience muscle stiffness or rigidity?"),
-    gr.Checkbox(label="Do you have issues with walking (e.g., shuffling, freezing)?"),
+    gr.Radio(["Yes", "No"], label="Do you experience tremors at rest?"),
+    gr.Radio(["Yes", "No"], label="Do you experience muscle stiffness or rigidity?"),
+    gr.Radio(["Yes", "No"], label="Do you have issues with walking (e.g., shuffling, freezing)?"),
     gr.Number(label="Age", value=50),
     gr.Audio(type="numpy", label="Upload Voice Recording (.wav, .mp3)")
 ]
 
+# Changed gr.Label to gr.Textbox to be safe
 outputs = [
-    gr.Textbox(label="Final Diagnosis"), # <--- Changed to Textbox
+    gr.Textbox(label="Final Diagnosis"),
     gr.Textbox(label="Raw Model Prediction"),
     gr.Textbox(label="Confidence Score")
 ]
@@ -82,4 +79,4 @@ demo = gr.Interface(
 )
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch() # No share=True needed for Hugging Face