Cross Validation problem about sparsevar HOT 1 CLOSED

Hi,
one of the parameter of the fitVAR function is method. This can be equal to "cv" or "timeSlice". The second method is the one you probably are referring to and should use if you think that cross validation is not appropriate for time series data. Hope this answer your question.

By the way: I believe I've read in some article (but I cannot remember which) that one can use also cross validation for time series (one introduces an error that in the end should be "negligible"). Sorry if I cannot help you more on this point, but currently I'm not working anymore in the field.

Best,
svazzole

from sparsevar.