This is a bug. Thanks for reporting it. PSPP is supposed to at least provide a header between sets of results when SPLIT FILE is in use, although some procedures can do better than that. This might be something that was introduced recently when I redid the output system, or it might be older. Either way, I've put it on my list of bugs to fix.
I have not yet heard back from anybody on whether the GUI fix I pushed made the crashes go away in builds after gc5cf5c. (I don't even know whether there has been a new Windows build released.) I'd like to hear about that, either a confirmation that they're gone or that they persist. On Tue, Jan 14, 2020 at 11:11 PM Dr. Oliver Walter <o.wal...@psychometrie-online.de> wrote: > > Dear Sir or Madam, > > I came across another problem in test version pspp 1.3.0-gc5cf5c on > Windows 10: In addition to permanent crashes this version does not show > which results of a linear regression belong to which group if you split > the data set into different groups, e.g. gender. Hence, I do not know > which results are for males and which are for females. I attached a jpg > to this e-mail. > > Kind regards, > > Dr. Oliver Walter > > _______________________________________________ > Bug-gnu-pspp mailing list > bug-gnu-p...@gnu.org > https://lists.gnu.org/mailman/listinfo/bug-gnu-pspp
