Update app.py

app.py

@@ -173,13 +173,19 @@ class CryptoAnalyzer:
             model.eval()
             with torch.no_grad():
                 predictions, confidence = model(X)
-                predictions = predictions.cpu().numpy().reshape(-1, 1)
-                predictions = self.price_scaler.inverse_transform(predictions).flatten()
-                y_np = y.cpu().numpy().reshape(-1, 1)
-                actual_prices = self.price_scaler.inverse_transform(y_np).flatten()
+            
+                # Fixing predictions and actual prices reshaping
+                predictions_reshaped = predictions.cpu().numpy().reshape(-1, 1)
+                predictions = self.price_scaler.inverse_transform(predictions_reshaped).flatten()
+
+                y_np_reshaped = y.cpu().numpy().reshape(-1, 1)
+                actual_prices = self.price_scaler.inverse_transform(y_np_reshaped).flatten()
+
+                # Metrics calculations
                 rmse = float(np.sqrt(np.mean((actual_prices - predictions) ** 2)))
                 mape = float(np.mean(np.abs((actual_prices - predictions) / actual_prices)) * 100)
                 r2 = float(1 - np.sum((actual_prices - predictions) ** 2) / np.sum((actual_prices - actual_prices.mean()) ** 2))
+            
                 dates = df.index[lookback:].strftime('%Y-%m-%d').tolist()
                 return {
                     'dates': dates,
@@ -198,6 +204,7 @@ class CryptoAnalyzer:
         except Exception as e:
             raise ValueError(f"Prediction failed: {str(e)}")
 
+
 def create_analysis_plots(symbol, days=180, lookback=30):
     try:
         analyzer = CryptoAnalyzer()