iode.Variables.trend_correction

Variables.trend_correction(input_file: str, lambda_: float, series: str | List[str] = None, log: bool = False)[source]

Implementation of the Hodrick-Prescott method for trend series (= variables) construction. The principle is the same as for deseasonalization: series read from a file are imported and transformed simultaneously.

If the log flag is set, the series are log-transformed before calculation and exp-transformed after calculation. The series values must therefore be strictly positive.

Parameters:

input_file: str: filepath to the input file.
lambda_: float: Lambda parameter of the Hodrick-Prescott method.
series: str or list(str), optional: list of series on which to apply the trend correction. Defaults to None (all variables).
log: bool, optional: Whether or not the series are log-transformed before calculation and exp-transformed after calculation. The series values must therefore be strictly positive. Defaults to False.