CalibrationResult._summary_table#

CalibrationResult._summary_table()[source]#

Build the structured summary: one row per series.

The PIT mean and standard deviation against their uniform benchmarks of 0.5 and \(1/\sqrt{12} \approx 0.289\), and the Berkowitz p-value; the notes give the reading of each departure and restate the one-step caveat.

Return type:

SummaryTable

Returns:

The SummaryTable that summary(), str(), and the notebook renderer display.

Example

>>> import numpy as np
>>> rng = np.random.default_rng(0)
>>> paths, outcomes = rng.standard_normal((80, 500, 2)), rng.standard_normal((80, 2))
>>> record = Calibration(paths, outcomes)
>>> table = record.compute()._summary_table()
>>> table.columns, table.rows[0]
(('series', 'PIT mean', 'PIT sd', 'Berkowitz p'), ('y1', '0.477', '0.285', '0.9890'))