Update utils.py

utils.py

@@ -6,7 +6,6 @@ from datetime import datetime, timedelta
 import plotly.graph_objects as go
 from plotly.subplots import make_subplots
 import spaces
-# Gunakan Chronos2Pipeline karena ini adalah kelas generik untuk model Chronos
 from chronos import Chronos2Pipeline
 
 def get_indonesian_stocks():
@@ -158,25 +157,25 @@ def format_large_number(num):
 @spaces.GPU(duration=120)
 def predict_prices(data, model=None, tokenizer=None, prediction_days=30):
     try:
-        # PENTING: Ganti ke Chronos-T5-Tiny untuk mencegah TimeOut/Crash GPU
-        pipeline = Chronos2Pipeline.from_pretrained("amazon/chronos-t5-tiny", device_map="auto")
+        # Panggil pipeline di sini untuk memastikan instance baru tiap run (mencegah error memori/state)
+        pipeline = Chronos2Pipeline.from_pretrained("amazon/chronos-2", device_map="auto")
         
+        # Chronos-2 with Covariate: Menggunakan Close (target) dan Volume (covariate)
         context_df = data[['Close', 'Volume']].reset_index()
         context_df.columns = ['timestamp', 'target', 'volume'] 
         context_df['id'] = 'stock_price' 
 
+        # Fix Error: Could not infer frequency & FIX VOLUME COVARIATE IMPUTATION
         context_df['timestamp'] = pd.to_datetime(context_df['timestamp'])
         context_df = context_df.set_index('timestamp').asfreq('D')
         
-        # IMPUTATION: Harga (target) ffill, Volume (covariate) fillna(0)
+        # IMPUTATION FIX: Target ffill, Covariate (Volume) fillna(0)
         context_df['target'] = context_df['target'].fillna(method='ffill')
         context_df['volume'] = context_df['volume'].fillna(0) 
         
-        # FINAL GUARD: Drop baris di awal yang mungkin masih NaN (jika history start pada hari non-dagang)
-        context_df = context_df.dropna(subset=['target', 'volume'])
-        
         context_df = context_df.reset_index()
         
+        # Pastikan kolom sesuai urutan Chronos-2: timestamp, target, covariate(s), id
         context_df['id'] = 'stock_price'
         context_df = context_df[['timestamp', 'target', 'volume', 'id']] 
 
@@ -190,11 +189,15 @@ def predict_prices(data, model=None, tokenizer=None, prediction_days=30):
                 quantile_levels=[0.1, 0.5, 0.9] 
             )
 
+        # --- FIX UTAMA: Pengecekan kolom hasil prediksi yang lebih ketat ---
         required_cols = ['target_0.1', 'target_0.5', 'target_0.9']
         if pred_df.empty or not all(col in pred_df.columns for col in required_cols):
+             # Jika gagal, pastikan kita tahu errornya dan melempar Runtime yang akan ditangkap di luar
              missing = [col for col in required_cols if col not in pred_df.columns]
              raise RuntimeError(f"Prediction failed. Result DataFrame is empty or incomplete. Missing: {missing}")
+        # ------------------------------------------------------------------
 
+        # Ekstraksi hasil prediksi kuantil
         q05_forecast = pred_df['target_0.5'].values.astype(np.float32)
         q09_forecast = pred_df['target_0.9'].values.astype(np.float32)
         q01_forecast = pred_df['target_0.1'].values.astype(np.float32)
@@ -216,12 +219,13 @@ def predict_prices(data, model=None, tokenizer=None, prediction_days=30):
             'change_pct': change_pct, 
             'q01': q01_forecast,
             'q09': q09_forecast,
-            'summary': f"AI Model: Amazon Chronos-T5-Tiny (Volume Covariate)\nPredicted High: {predicted_high:.2f}\nPredicted Low: {predicted_low:.2f}\nExpected Change: {change_pct:.2f}%"
+            'summary': f"AI Model: Amazon Chronos-2 (Volume Covariate)\nPredicted High: {predicted_high:.2f}\nPredicted Low: {predicted_low:.2f}\nExpected Change: {change_pct:.2f}%"
         }
         
     except Exception as e:
         error_message = f'Model prediction failed: {e}'
         print(f"Error in prediction: {e}")
+        # Mengembalikan objek error yang valid
         return {'values': [], 'dates': [], 'high_30d': 0, 'low_30d': 0, 'mean_30d': 0, 'change_pct': 0, 'summary': error_message, 'q01': [], 'q09': []}
 
 def create_prediction_chart(data, predictions):
@@ -230,8 +234,10 @@ def create_prediction_chart(data, predictions):
         
     fig = go.Figure()
     
+    # Historical Price: Menggunakan seluruh data historis
     fig.add_trace(go.Scatter(x=data.index, y=data['Close'].values, name='Historical Price', line=dict(color='blue', width=2)))
 
+    # Prediction Interval (Band): Menggunakan Q0.1 dan Q0.9
     fig.add_trace(go.Scatter(
         x=predictions['dates'], 
         y=predictions['q09'], 
@@ -248,10 +254,11 @@ def create_prediction_chart(data, predictions):
         fillcolor='rgba(255,182,193,0.3)'
     ))
     
+    # Median Forecast (Q0.5) - Garis Utama Prediksi
     fig.add_trace(go.Scatter(x=predictions['dates'], y=predictions['values'], name='Median Forecast (Q0.5)', line=dict(color='red', width=3, dash='solid')))
 
     fig.update_layout(
-        title=f'Probabilistic Price Forecast - Next {len(predictions["dates"])} Days (Chronos-T5-Tiny)', 
+        title=f'Probabilistic Price Forecast - Next {len(predictions["dates"])} Days (Chronos-2)', 
         xaxis_title='Date', 
         yaxis_title='Price (IDR)', 
         hovermode='x unified', 
@@ -259,6 +266,7 @@ def create_prediction_chart(data, predictions):
         legend=dict(yanchor="top", y=0.99, xanchor="left", x=0.01)
     )
     
+    # Menandai titik harga terakhir
     last_hist_date = data.index[-1]
     last_hist_price = data['Close'].iloc[-1]
     fig.add_trace(go.Scatter(