Update utils.py

utils.py

@@ -7,7 +7,6 @@ import plotly.graph_objects as go
 import plotly.express as px
 from plotly.subplots import make_subplots
 import spaces
-from chronos import BaseChronosPipeline
 
 def get_indonesian_stocks():
     return {
@@ -217,30 +216,39 @@ def format_large_number(num):
 def predict_prices(data, model=None, tokenizer=None, prediction_days=30):
     try:
         prices = data['Close'].values.astype(np.float32)
+        try:
+            from chronos import BaseChronosPipeline
+        except Exception as ie:
+            return {
+                'values': [],
+                'dates': [],
+                'high_30d': 0,
+                'low_30d': 0,
+                'mean_30d': 0,
+                'change_pct': 0,
+                'summary': 'chronos package not installed. install with: pip install chronos-forecasting'
+            }
         pipeline = BaseChronosPipeline.from_pretrained("amazon/chronos-bolt-base", device_map="auto")
         with torch.no_grad():
             forecast = pipeline.predict(context=torch.tensor(prices), prediction_length=prediction_days)
-        mean_forecast = forecast.mean(dim=1).squeeze().cpu().numpy()
+        if hasattr(forecast, 'mean'):
+            mean_forecast = forecast.mean(dim=1).squeeze().cpu().numpy()
+        else:
+            mean_forecast = forecast.mean(dim=1).squeeze().cpu().numpy()
         pred_len = len(mean_forecast)
         last_price = prices[-1]
-        predicted_high = np.max(mean_forecast)
-        predicted_low = np.min(mean_forecast)
-        predicted_mean = np.mean(mean_forecast)
-        change_pct = ((predicted_mean - last_price) / last_price) * 100
+        predicted_high = np.max(mean_forecast) if pred_len > 0 else 0
+        predicted_low = np.min(mean_forecast) if pred_len > 0 else 0
+        predicted_mean = np.mean(mean_forecast) if pred_len > 0 else 0
+        change_pct = ((predicted_mean - last_price) / last_price) * 100 if last_price != 0 and pred_len > 0 else 0
         return {
             'values': mean_forecast,
-            'dates': pd.date_range(start=data.index[-1] + timedelta(days=1), periods=pred_len, freq='D'),
+            'dates': pd.date_range(start=data.index[-1] + timedelta(days=1), periods=pred_len, freq='D') if pred_len > 0 else [],
             'high_30d': predicted_high,
             'low_30d': predicted_low,
             'mean_30d': predicted_mean,
             'change_pct': change_pct,
-            'summary': f"""
-AI Model: Amazon Chronos-Bolt (Base)
-Prediction Period: {pred_len} days
-Expected Change: {change_pct:.2f}%
-Confidence: Medium
-Note: AI predictions are for reference only and not financial advice
-            """.strip()
+            'summary': f"AI Model: Amazon Chronos-Bolt (Base)\nPrediction Period: {pred_len} days\nExpected Change: {change_pct:.2f}%\nConfidence: Medium\nNote: AI predictions are for reference only and not financial advice"
         }
     except Exception as e:
         print(f"Error in prediction: {e}")