get_metric_summaries

rubin_sim.maf.run_comparison.get_metric_summaries(run_families=(), metric_sets=(), runs=(), metrics=(), summary_source=None, runs_source=None, metric_set_source=None, run_order='family', metric_order='summary')

Get summary metric values for a set of runs and metrics.

Parameters:

run_familiesiterable [str]

Families of runs to include in the summary.

metric_setsiterable [str]

Sets of metrics to include in the summary.

runsiterable [str]

Runs to include in the summary (in addition to any that are part of families included in run_families).

metricsiterable [str]

Metrics to include in the summary (in addition to any that are part of sets included in metric_sets).

summary_sourcestr or pandas.DataFrame

File name or URL for the file from which to load the data. If it is set to None, the data is loaded from the URL specified by the archive.METRIC_SET_SOURCE constant. If the supplied value is a pandas.DataFrame, it the table returned will be a subset of this supplied table.

run_sourcepandas.DataFrame or str

Either a pandas.DataFrame of runs metadata (as returned by archive.get_runs), or a file name or URL for the json file from which to load the run metadata. If it is set to None, the data is loaded from the URL specified by the archive.RUNS_SOURCE constant.

metric_set_sourcepandas.DataFrame or str

Either a pandas.DataFrame of metric set specifications (as returned by archive.get_metric_sets) or a file name or URL for the json file from which to load the data. If it is set to None, the data is loaded from the URL specified by the archive.SUMMARY_SOURCE constant.

run_orderstr

Sort runs according to family definition (“family”) or summary file (“summary”) order.

metric_orderstr

Sort metrics according to set definition (“set”) or summary file (“summary”) order.

Returns:

summariespandas.DataFrame

Metric summary values are returned in a pandas.DataFrame, with each column providing the metrics for one run, and each row the values for one metric. The metric names constitute the index, and the column names are the canonical run names.