roastcoffea.aggregation.core

Core metrics aggregator combining all aggregation modules.

Classes

MetricsAggregator(backend)

Main aggregator combining workflow, worker, and efficiency metrics.
class roastcoffea.aggregation.core.MetricsAggregator(backend)[source]

Bases: object

Main aggregator combining workflow, worker, and efficiency metrics.

Parameters:

backend (str)

aggregate(coffea_report, tracking_data, t_start, t_end, custom_metrics=None, span_metrics=None, processor_name=None, chunk_metrics=None, section_metrics=None)[source]

Aggregate all metrics from workflow run.

Parameters:

  • coffea_report (dict) – Coffea report

  • tracking_data (dict, optional) – Backend tracking data

  • t_start (float) – Start time

  • t_end (float) – End time

  • custom_metrics (dict, optional) – Per-dataset metrics

  • span_metrics (dict, optional) – Dask Spans cumulative_worker_metrics

  • processor_name (str, optional) – Name of processor class for filtering fine metrics

  • chunk_metrics (list of dict, optional) – Per-chunk metrics from @track_metrics decorator

  • section_metrics (list of dict, optional) – Section metrics from track_section() and track_memory()

Returns:

Combined metrics

Return type:

dict