Am 31.01.2024 um 11:47 schrieb d.lalountas(a)minfin.gr: If I understand your problem correctly, then sdiff should still be OK for you, even in the MIDAS case. For example, in the gdp_midas.gdt example dataset that comes with gretl, industrial production is a monthly variable, and monthly is the "high" frequency, so it comes as three Midas-series, namely indpro_m1, indpro_m2, indpro_m3. But the "core" or "low" frequency of the dataset is quarterly. So if you apply sdiff() to, say, indpro_m1, then the obs for, say, January 2010 is embedded in that series in the 2010Q1 low-freq period. Going back 4 quarters then gives you exactly what you want, January 2009 within 2009Q1. So: <hansl> open gdp_midas.gdt list L = indpro* list direct = sdiff(L) # manual comparison series indirectM1 = indpro_m1 - indpro_m1(-$pd) # $pd is 4 here # check the equality: eval sum(abs(sd_indpro_m1 - indirectM1)) </hansl> However, it seems that you have to re-define the result from sdiff() (the "direct" list here) explicitly as a MIDAS list again (setinfo direct --midas), for example for plotting purposes, which is a slight nuisance. cheers sven

Am 02.02.2024 um 08:30 schrieb d.lalountas(a)minfin.gr: If I understand you correctly, then you should be OK with using sdiffin the hi-freq context. In a quarterly and monthly context, seasonal always means one year ago (or $pd periods, where $pd==4 for quarterly and $pd==12 for monthly), so that matches. The same should hold for weekly data, I guess, but subject to testing and verification. For daily data, however, things are likely to become trickier. There, the standard meaning of seasonal is (I believe) from week to week, and there you then have a mismatch with the seasonal definition for weekly data. Also, all this assumes that there's just one layer of seasonality. For monthly data, in principle you could think of both a quarterly and an annual cycle. But so far gretl does not natively support several coexisting seasonalities, AFAIK. cheers sven