kats.utils.decomposition module¶

class kats.utils.decomposition.TimeSeriesDecomposition(data: kats.consts.TimeSeriesData, decomposition='additive', method='STL', **kwargs)[source]¶

Bases: object

Model class for Time Series Decomposition.

This class provides utilities to decompose an input time series

data¶

the input time series data as TimeSeriesData

decomposition¶

additive or multiplicative decomposition

method¶

STL decompostion or seasonal_decompose

Specific arguments to seasonal_decompose and STL functions can be passed via kwargs

decomposer()[source]¶

Decompose the time series.

Parameters

None. –

Returns

trend : Trend seasonal : Seasonality, and rem : Residual

Return type

A dictionary with three time series for the three components

plot()[source]¶

Plot the original time series and the three decomposed components.