MultiErrorReport

from tmlt.tune import MultiErrorReport

class tmlt.tune.MultiErrorReport(reports)

Bases: object

Output of an error report run across multiple input combinations.

This class is not intended to be constructed directly. Instead, it is returned by the multi_error_report() method.

property reports: List[ErrorReport]

Return the error reports.

__iter__()

Return an iterator over the error reports.

Return type:

Iterator[ErrorReport]

dataframes()

Return the result dataframes for each metric combined across runs.

Each metric produces a separate DataFrame. The DataFrames are keyed by the result column names.

Return type:

Dict[str, DataFrame]

dataframe()

Return a DataFrame representation of the error reports.

The DataFrame will have a column for each parameter and metric. If all metrics have the same grouping columns, the DataFrame will include these grouping columns. If not all metrics have the same (or no) grouping column, this method will throw an error.

If some combinations of grouping columns values are associated with only some of the metrics, the missing metrics appear as null values in the output DataFrame.

Return type:

DataFrame