Danger

You are looking at the documentation for an older version of the SDV! We are no longer supporting or maintaining this version of the software

Click here to go to the new docs pages.

sdv.evaluation.evaluate

sdv.evaluation.evaluate(synthetic_data, real_data=None, metadata=None, root_path=None, table_name=None, metrics=None, aggregate=True)[source]

Apply multiple metrics at once.

Parameters

  • synthetic_data (dict[str, pandas.DataFrame] or pandas.DataFrame) – Map of names and tables of synthesized data. When evaluating a single table, a single pandas.DataFrame can be passed alone.

  • real_data (dict[str, pandas.DataFrame] or pandas.DataFrame) – Map of names and tables of real data. When evaluating a single table, a single pandas.DataFrame can be passed alone. If metadata is None, this parameter must be a dataframe.

  • metadata (str, dict, Metadata or None) – Metadata instance or details needed to build it.

  • root_path (str) – Relative path to find the metadata.json file when needed.

  • metrics (list[str]) – List of metric names to apply.

  • table_name (str) – Table name to be evaluated, only used when synthetic_data is a pandas.DataFrame and real_data is None.

  • aggregate (bool) – If get_report is False, whether to compute the mean of all the normalized scores to return a single float value or return a dict containing the score that each metric obtained. Defaults to True.

Returns

float or sdmetrics.MetricsReport

sdv.evaluation sdv.demo