Re: [Gretl-devel] autoregressive "genr" bug

OK, I will defer on that. Sorry for the noise. I suppose, then, that the bug exposed by the panel case that I mentioned is actually that this treatment is not granted to the first observation for each individual, in case of an autoregressive "genr". (I can see no rationale for granting it just for the first individual, which is what happens now.)