[Gretl-users] Re: GRETL package for testing fractional integration with trends and breaks
Am 02.07.2021 um 14:29 schrieb Artur T.:
Hi,
I am not aware of any functionality of user-contributed package
covering what you ask fo.
Perhaps the "gen_hurst" contributed package by Joachim Reinhardt might
be relevant. According to its description it offers detrended
fluctuation analysis (DFA) as an option to estimate the long-range
dependence parameter. Don't know if that's what Taylan had in mind.
(Apart from the simple alternative to pre-adjust any time series for
trend components and shifts, with the obvious complications for
subsequent inference.)
cheers
sven