Hello,
For a "full" analysis, if the named output folder already exists, FEAT will create a new folder, with an extra "+" in the name, if a single analysis is creating multiple folders at the same time, it could possibly be the cluster restarting failed jobs ( if it has that capability ).
Kind Regards
Matthew
> Hi FSL Users -
>
> I am currently using FEAT to regress out neural signal associated with CSF and WM. I'm using a computer cluster (which may be related to the problem... but maybe not) - and using command line bash scripts to call FEAT. However, i'm finding that FEAT is creating many many folders... some of which finish the GLM to completion. I'm a little hesitant to trust the data - and i'm concerned about the proliferation of folders.
>
> For example, I will start the process and FSL will produce
>
> SUBJXXX.feat
> SUBJXXX+.feat
> SUBJXXX++.feat
> SUBJXXX+++.feat
> SUBJXXX++++.feat and so forth.
>
> Usually the stats finish in the original folder - but occasionally in the second folder. Is there any particular reason anyone can think of that would cause this?
>
> Thanks,
> Tracy
|