Fix model loading and prediction issues - working locally

app.py

@@ -18,19 +18,65 @@ def utility_processor():
         return IMAGE_BASE_URL + path
     return dict(get_image_url=get_image_url, images=IMAGES)
 
-# Update model loading with error handling
+# Load and prepare data for scalers
 try:
-    model = pickle.load(open('model.pkl', 'rb'))
-    sc = pickle.load(open('standscaler.pkl', 'rb'))
-    ms = pickle.load(open('minmaxscaler.pkl', 'rb'))
+    # Try to load the models first
+    with open('model.pkl', 'rb') as f:
+        model = pickle.load(f)
+    with open('standscaler.pkl', 'rb') as f:
+        sc = pickle.load(f)
+    with open('minmaxscaler.pkl', 'rb') as f:
+        ms = pickle.load(f)
+    print("Models loaded successfully")
 except Exception as e:
     print(f"Error loading models: {e}")
-    # Provide default/dummy models if loading fails
-    from sklearn.ensemble import RandomForestClassifier
-    model = RandomForestClassifier()
-    from sklearn.preprocessing import StandardScaler, MinMaxScaler
-    sc = StandardScaler()
-    ms = MinMaxScaler()
+    try:
+        # Load the dataset and fit scalers
+        data = pd.read_csv('Crop_recommendation.csv')
+        features = ['N', 'P', 'K', 'temperature', 'humidity', 'ph', 'rainfall']
+        X = data[features].values
+        y = data['label'].values
+
+        # Create and fit scalers with actual data
+        from sklearn.preprocessing import StandardScaler, MinMaxScaler
+        ms = MinMaxScaler()
+        ms.fit(X)
+        
+        # Transform with MinMaxScaler first
+        X_minmax = ms.transform(X)
+        
+        sc = StandardScaler()
+        sc.fit(X_minmax)
+        
+        # Create and fit model with transformed data
+        from sklearn.ensemble import RandomForestClassifier
+        model = RandomForestClassifier(n_estimators=100, random_state=42)
+        model.fit(sc.transform(X_minmax), y)
+        print("Models fitted successfully with dataset")
+        
+        # Save the fitted models
+        with open('model.pkl', 'wb') as f:
+            pickle.dump(model, f)
+        with open('standscaler.pkl', 'wb') as f:
+            pickle.dump(sc, f)
+        with open('minmaxscaler.pkl', 'wb') as f:
+            pickle.dump(ms, f)
+            
+    except Exception as e:
+        print(f"Error fitting models: {e}")
+        # Create and fit default scalers with sample data as fallback
+        sample_data = np.array([[90, 40, 40, 20, 80, 7, 200],
+                              [20, 30, 10, 25, 60, 6, 100]])
+        
+        ms = MinMaxScaler()
+        ms.fit(sample_data)
+        
+        sc = StandardScaler()
+        sc.fit(sample_data)
+        
+        model = RandomForestClassifier()
+        model.fit(sample_data, [1, 2])
+        print("Using default models")
 
 # Define routes
 @app.route('/')
@@ -48,6 +94,7 @@ def recommendation():
 
     if request.method == 'POST':
         try:
+            # Get form data
             N = float(request.form['Nitrogen'])
             P = float(request.form['Phosporus'])
             K = float(request.form['Potassium'])
@@ -56,27 +103,41 @@ def recommendation():
             ph = float(request.form['Ph'])
             rainfall = float(request.form['Rainfall'])
 
+            # Prepare features
             feature_list = [N, P, K, temp, humidity, ph, rainfall]
             single_pred = np.array(feature_list).reshape(1, -1)
 
-            scaled_features = ms.transform(single_pred)
-            final_features = sc.transform(scaled_features)
-            prediction = model.predict(final_features)
-
-            crop_dict = {1: "Rice", 2: "Maize", 3: "Jute", 4: "Cotton", 5: "Coconut", 6: "Papaya", 7: "Orange",
-                        8: "Apple", 9: "Muskmelon", 10: "Watermelon", 11: "Grapes", 12: "Mango", 13: "Banana",
-                        14: "Pomegranate", 15: "Lentil", 16: "Blackgram", 17: "Mungbean", 18: "Mothbeans",
-                        19: "Pigeonpeas", 20: "Kidneybeans", 21: "Chickpea", 22: "Coffee"}
+            # Scale features
+            try:
+                scaled_features = ms.transform(single_pred)
+                final_features = sc.transform(scaled_features)
+                
+                # Make prediction
+                prediction = model.predict(final_features)
+                
+                # Map prediction to crop name
+                crop_dict = {1: "Rice", 2: "Maize", 3: "Jute", 4: "Cotton", 5: "Coconut", 6: "Papaya", 7: "Orange",
+                            8: "Apple", 9: "Muskmelon", 10: "Watermelon", 11: "Grapes", 12: "Mango", 13: "Banana",
+                            14: "Pomegranate", 15: "Lentil", 16: "Blackgram", 17: "Mungbean", 18: "Mothbeans",
+                            19: "Pigeonpeas", 20: "Kidneybeans", 21: "Chickpea", 22: "Coffee"}
 
-            if prediction[0] in crop_dict:
-                crop = crop_dict[prediction[0]]
-                result = "{} is the best crop to be cultivated right there".format(crop)
-            else:
-                result = "Sorry, we could not determine the best crop to be cultivated with the provided data."
+                if prediction[0] in crop_dict:
+                    crop = crop_dict[prediction[0]]
+                    result = "{} is the best crop to be cultivated right there".format(crop)
+                else:
+                    result = "Sorry, we could not determine the best crop to be cultivated with the provided data."
+                    
+            except Exception as e:
+                print(f"Error in prediction: {e}")
+                result = "An error occurred during prediction. Please try again."
+                
         except Exception as e:
+            print(f"Error processing form data: {e}")
             result = f"An error occurred: {str(e)}"
 
-    return render_template('recommendation.html', result=result, N=N, P=P, K=K, temp=temp, humidity=humidity, ph=ph, rainfall=rainfall)
+    return render_template('recommendation.html', result=result, 
+                         N=N, P=P, K=K, temp=temp, 
+                         humidity=humidity, ph=ph, rainfall=rainfall)
 
 if __name__ == "__main__":
     app.run(host="0.0.0.0", port=7860)