Fix historical data not updating after automatic reloads (#9)

- fix historical data updates (30401c0a62ad4d0ca69e642f4ab5c0faf2c23e27)


Co-authored-by: Abdennacer Badaoui <badaoui@users.noreply.huggingface.co>

data.py

@@ -560,6 +560,13 @@ class CIResults:
         # Load all historical data at startup
         self.load_all_historical_data()
         
+        # Update historical_df with latest available dates after reload
+        if self.available_dates:
+            start_date_val = self.available_dates[-1]  # Last date (oldest)
+            end_date_val = self.available_dates[0]       # First date (newest)
+            self.load_historical_data(start_date_val, end_date_val)
+            logger.info(f"Updated historical_df with {len(self.historical_df)} records")
+        
         # Log and return distant load status
         logger.info(f"Data loaded successfully: {len(self.available_models)} models")
         logger.info(f"Models: {self.available_models[:5]}{'...' if len(self.available_models) > 5 else ''}")
@@ -578,7 +585,7 @@ class CIResults:
         logger.info(json.dumps(msg, indent=4))
 
     def load_all_historical_data(self) -> None:
-        """Load all available historical data at startup."""
+        """Load all available historical data. Replaces existing data to ensure latest dates are included."""
         try:
             if not self.available_dates:
                 logger.warning("No available dates found, skipping historical data load")
@@ -587,8 +594,12 @@ class CIResults:
             
             logger.info(f"Loading all historical data for {len(self.available_dates)} dates...")
             start_date, end_date = self.available_dates[-1], self.available_dates[0]
+            logger.info(f"Date range: {start_date} to {end_date}")
             self.all_historical_data = get_historical_data(start_date, end_date, self.sample_data)
             logger.info(f"All historical data loaded: {len(self.all_historical_data)} records")
+            if not self.all_historical_data.empty:
+                unique_dates = sorted(self.all_historical_data['date'].unique())
+                logger.info(f"Loaded dates: {unique_dates[0]} to {unique_dates[-1]} ({len(unique_dates)} unique dates)")
         except Exception as e:
             logger.error(f"Error loading all historical data: {e}")
             self.all_historical_data = pd.DataFrame()
@@ -639,3 +650,4 @@ class CIResults:
         timer.daemon = True
         timer.start()
         logger.info("Data auto-reload scheduled every 15 minutes")
+