iode.Variables.seasonal_adjustment

Variables.seasonal_adjustment(input_file: str, eps_test: float = 5.0, series: str | List[str] = None)[source]

Eliminate seasonal variations in monthly series (= variables).

The method constructs deseasonalized series using the Census XI method, as well as cyclical and stochastic trend components. Thus, In addition to the deseasonalized series (named after the series in the input file), the result is two other series:

the series containing the cyclical trend component named : _C<name> where name is the original name

the series containing the stochastic component called : _I<name> where name is the original name

Note that the deseasonalized series is the product of the other two.

To do this, the list of series (= variables) are loaded from the specified file input_file into the current IODE Variables database and simultaneously modifies the selected series if necessary.

The new series are added to or replace (for existing names) those in the current IODE Variables database.

In the case of non-existent value (NA) for one of the periods, the result is NA for the whole year.

Parameters:

input_file: str or Path: Filepath to the input file.
eps_test: float, optional: Criterion verifying whether a seasonal influence is present in a series Default to 5.0.
series: str or list(str), optional: list of series on which to apply the seasonal adjustment. Defaults to None (all variables).