plot_comparison#

er_evaluation.plots.plot_comparison(predictions, metrics={'Pairwise F1': <function pairwise_f>, 'Pairwise precision': <function pairwise_precision>}, color_continuous_scale='Blues', **kwargs)[source]#

Plot metrics computed for all prediction pairs.

This is meant to help investigate the similarity and differences between a set of disambiguations.

Parameters:

  • predictions (dict) – dictionary of membership vectors to compare.

  • metrics (dict, optional) – Dictionary of metrics to compute. Defaults to DEFAULT_COMPARISON_METRICS.

  • **kwargs (optional) – Additional arguments to pass to plotly express for plot creation.

Examples

>>> from er_evaluation.datasets import load_pv_disambiguations
>>> predictions, _ = load_pv_disambiguations()
>>> fig = plot_comparison(predictions)
>>> fig.show()