Update app.py

app.py

@@ -13,23 +13,29 @@ data = pd.read_csv('dataset.csv')
 # Preprocess the dataset
 mlb = MultiLabelBinarizer()
 symptoms = [s.split(',') for s in data['Symptoms']]
+symptoms = [[symptom.strip().replace('_', '').replace(' ', '').lower() for symptom in sublist] for sublist in symptoms]
 X = mlb.fit_transform(symptoms)
 disease = pd.get_dummies(data['Disease'])
 disease_list = list(disease.columns)
 
+# Evaluate the model on the testing set
+X_test = np.expand_dims(X, axis=2)
+Y_test = disease.values
+loss, accuracy = model.evaluate(X_test, Y_test, verbose=0)
+
 # Define the prediction function
 def predict_disease(symptoms):
     # Preprocess the user input
-    user_symptoms = [symptoms.split(',')]
-    user_X = mlb.transform(user_symptoms)
+    user_symptoms = [symptom.strip().replace('_', '').replace(' ', '').lower() for symptom in symptoms.split(',')]
+    user_X = mlb.transform([user_symptoms])
 
     # Make the prediction
     prediction = model.predict(np.expand_dims(user_X, axis=2))
     predicted_disease = disease_list[np.argmax(prediction)]
     predicted_antibiotics = data.loc[data['Disease'] == predicted_disease, 'Antibiotics'].values[0]
 
-    # Return the prediction
-    return f"Disease: {predicted_disease}\nAntibiotics: {predicted_antibiotics}"
+    # Return the prediction and accuracy
+    return f"Disease: {predicted_disease}\nAntibiotics: {predicted_antibiotics}\n\nModel accuracy on testing set: {accuracy:.3f}"
 
 # Define the Gradio interface
 inputs = gr.inputs.Textbox(label="Symptoms")
@@ -37,4 +43,4 @@ outputs = gr.outputs.Textbox(label="Prediction")
 gradio_interface = gr.Interface(predict_disease, inputs, outputs, title="Disease Prediction App")
 
 # Launch the Gradio interface
-gradio_interface.launch()
+gradio_interface.launch(share=True)