Fix filtering for components in CBPE result

Author: michael-nml

nannyml/performance_estimation/confidence_based/results.py

@@ -104,6 +104,7 @@ def _filter(self, period: str, metrics: Optional[List[str]] = None, *args, **kwa
         metric_column_names = [name for metric in filtered_metrics for name in metric.column_names]
 
         res = super()._filter(period, metric_column_names, *args, **kwargs)
+        res.metrics = filtered_metrics
 
         return res