v1 trial

NdR_disease.py

@@ -197,7 +197,7 @@ def run_disease_train():
 
 def get_user_input_and_predict(linear_model, neural_model, scaler, conditions, num_classes):
     st.write("\nAdjust the sliders for the following symptoms on a scale from 0 (none) to 10 (severe):")
-    
+
     # Feature names
     feature_names = ['Fever', 'Cough', 'Sneezing', 'Runny Nose', 'Nausea', 'Vomiting', 
                      'Diarrhea', 'Headache', 'Fatigue', 'Stress Level']
@@ -205,24 +205,32 @@ def get_user_input_and_predict(linear_model, neural_model, scaler, conditions, n
     # Create sliders for user input
     user_features = []
     for feature in feature_names:
-        value = st.slider(feature, 0, 10, 5)  # Default value set to 5
-        user_features.append(value)
-    
+        # Use session state to retain slider values
+        default_value = 5
+        user_input = st.slider(feature, 0, 10, st.session_state.get(feature, default_value))
+        st.session_state[feature] = user_input
+        user_features.append(user_input)
+
     # Calculate interaction terms
     interaction_term = np.sin(user_features[7]) * np.log1p(user_features[9])  # Headache and Stress Level
     interaction_term2 = user_features[0] * user_features[4]  # Fever * Nausea
     user_features.extend([interaction_term, interaction_term2])
-    
+
     # Add a button to trigger the predictions
     if st.button('Calculate Predictions'):
+        # Set session state to indicate the button was clicked
+        st.session_state.button_clicked = True
+
+    # Only show predictions if the button has been clicked
+    if st.session_state.button_clicked:
         # Normalize features
         user_features = scaler.transform([user_features])
         user_tensor = torch.from_numpy(user_features).float()
-        
+
         # Random prediction
         random_pred = np.random.randint(num_classes)
         st.write(f"\nRandom Prediction: {conditions[random_pred]}")
-        
+
         # Linear Model Prediction
         linear_model.eval()
         with torch.no_grad():
@@ -230,7 +238,7 @@ def get_user_input_and_predict(linear_model, neural_model, scaler, conditions, n
             _, predicted = torch.max(outputs.data, 1)
             linear_pred = predicted.item()
         st.write(f"Linear Model Prediction: {conditions[linear_pred]}")
-        
+
         # Neural Network Prediction
         neural_model.eval()
         with torch.no_grad():

app.py

@@ -10,6 +10,9 @@ from sklearn.model_selection import train_test_split
 
 from NdR_disease import run_disease_train, get_user_input_and_predict
 
+if 'button_clicked' not in st.session_state:
+    st.session_state.button_clicked = False
+
 
 # Function for male superhero task
 def run_male_superhero_task():