gretapy.ds.read_imaginary_metrics

gretapy.ds.read_imaginary_metrics(seed=None, remove_paired=True, verbose=False)

Read benchmark metrics for an imaginary method.

Calls read_metrics() and samples one row per unique benchmark configuration (class, task, db, organism, dataset), then sets the method name to 'ImaginaryMethod'. Useful for baseline comparisons or testing visualizations.

Parameters:

• seed (int or None) – Seed for numpy.random.default_rng(). Default is None (non-deterministic).

• remove_paired (bool (default: True)) – Passed to read_metrics(). Default is True.

• verbose (bool (default: False)) – Passed to read_metrics(). Default is False.

Return type:

DataFrame

Returns:

pd.DataFrame A DataFrame with the same columns as read_metrics(), with one row per unique benchmark configuration and name set to 'ImaginaryMethod'.