Hi,
Can you upload the full report_log.html file(s) from the problem analyses?
>
> I am also getting these problems (see the other thread I started
> yesterday), but:
>
> 1) When running FE, not ME
> 2) On the latest version of FSL (4.0.3-1)
> 3) for Ubuntu, not Debian
> 4) On IDENTICAL analyses which worked fine just a few days ago, before
> updating to the latest version (so, it can't be due to outliers in my
> data, else I should have had these errors last week, and i haven't made
> any changes to or noticed anything funny in my computer)
> 5) On FE analyses which seem to invoke FLAME 1+2 processes (I get errors
> telling me that the results of FLAME 1 and FLAME 1+2 are very different
> but I chose the FE option, which usually takes 10 minutes, but now takes
> hours)
>
>
> this is very much a shot-in-the dark, but, since there are quite a few
> posts in the archives on these sorts of topics, i wondered:
>
> Since the number-labelling of the four possible higher-level analyses in
> the design.fsf file in feat is:
> 1st option FE - 3
> 2nd OLS - 0
> 3rd ME(FLAME 1) - 2
> 4th ME(FLAME 1+2)- 1
>
> is it possible that, somewhere along the way, and differently in different
> versions, this non-intuitive number/ordering (3-0-2-1) is causing some
> confusions or human error in the programs?
>
> I can't otherwise think why when I select the first option FE (#3), I get
> error messages related to the fourth option (#1) on identical analyses
> which worked fine before the latest update...
>
> i hope we can find the problem(s) soon - i can no longer work...
>
> cheers,
> nick
>
>
|