Update components/predict.py

components/predict.py

@@ -1,20 +1,44 @@
-import gradio as gr
+# components/predict.py
+
 import pickle
 import numpy as np
 from utils.db import log_prediction
-from components.auth import user_session
 
-model = pickle.load(open("model.pkl", "rb"))
+# Load the machine learning model from the file
+try:
+    model = pickle.load(open("model.pkl", "rb"))
+except FileNotFoundError:
+    raise RuntimeError("The 'model.pkl' file was not found. Please run the 'create_dummy_model.py' script first.")
+
+
+def predict(pregnancies, glucose, blood_pressure, insulin, bmi, age, user_state):
+    """
+    Performs diabetes prediction based on user input.
+    """
+    # First, check if the user is logged in by checking the user_state
+    if not user_state.get("logged_in"):
+        return "❌ Please log in on the 'Login / Signup' tab before making a prediction."
+
+    # Validate inputs
+    if any(v is None for v in [pregnancies, glucose, blood_pressure, insulin, bmi, age]):
+        return "❌ Please fill in all the fields."
 
-def predict(pregnancies, glucose, blood_pressure, insulin, bmi, age):
-    if not user_session["email"]:
-        return "❌ Please log in first."
+    # Prepare the data for the model
+    input_data = [pregnancies, glucose, blood_pressure, insulin, bmi, age]
+    data_np = np.array(input_data).reshape(1, -1)
 
-    data = [pregnancies, glucose, blood_pressure, insulin, bmi, age]
-    data_np = np.array(data).reshape(1, -1)
+    # Make the prediction
+    try:
+        prediction = model.predict(data_np)[0]
+        result_text = "Diabetic" if prediction == 1 else "Not Diabetic"
+        
+        # Format the result with an emoji
+        final_result = f"✅ Result: {result_text}"
 
-    prediction = model.predict(data_np)[0]
-    result = "✅ Diabetic" if prediction == 1 else "🟢 Not Diabetic"
+        # Log the prediction to the database in the background
+        log_prediction(input_data, result_text, user_state["email"])
+        
+        return final_result
 
-    log_prediction(data, int(prediction), user_session["email"])
-    return result
+    except Exception as e:
+        return f"An error occurred during prediction: {e}"