DecompositionFilterResult._summary_table#

DecompositionFilterResult._summary_table()[source]#

Cycle scale per series, absolute and against the input’s growth, with the caveats.

One row per series: the cycle’s standard deviation and its ratio to the standard deviation of the reconstructed level’s first difference. The metadata block carries whichever of offset, penalty, cutoff, order, horizon and lags the method set; the notes come from _method_notes().

Return type:

SummaryTable

Returns:

The SummaryTable.

Example

>>> rng = np.random.default_rng(0)
>>> y = np.cumsum(rng.standard_normal(200))
>>> table = HodrickPrescottFilter().filter(y)._summary_table()
>>> table.title
'Hodrick-Prescott Trend-Cycle Decomposition'
>>> table.metadata[2:]
(('Penalty', '1600'), ('Cutoff period', '39.7'), ('Order', '2'))
>>> table.rows
(('series 1', '1.4437', '1.498'),)