Hi - it looks like it is failing to create training files because your hand label files are wrong - some include noise component numbers which are greater than the actual number of ICA components - for some runs ! Cheers, Steve. On 8 Nov 2014, at 15:48, David V. Smith <[log in to unmask]> wrote: > Thanks, Steve. I can’t find the old data upload URL, so here’s a Dropbox link to a FIX_debugging.tar.gz file containing all of the output and error files from my latest FIX training attempt. > > https://www.dropbox.com/s/h1o6uuvni72anvr/FIX_debugging.tar.gz > > I’ve copied individual features (from the subject.feat/fix folder) to the “training_features” subfolder. > > Thanks for taking the time to look at this. > > Cheers, > David > > > > On Nov 8, 2014, at 4:21 AM, Stephen Smith <[log in to unmask]> wrote: > >> Hi David - it's hard to tell from this what's going wrong with the training-weights creation. If you want to create a tarfile containing all of the output folders/files being created we can take a look. >> >> Regards, Steve. >> >> >> On 7 Nov 2014, at 19:42, David V. Smith <[log in to unmask]> wrote: >> >>> Hi Steve, >>> >>> My mistake — I wasn’t actually using the compiled files (I didn’t change the default). FSL_FIX_MATLAB_MODE=1 >>> >>> Unfortunately, I don’t have SGE set up here; I’m not sure that option is set up by default (it doesn’t seem to be). >>> >>> I tried FIX with the Standard.RData training weight files, and it worked fine (as did the subsequent cleaning step). So, it looks like I’m mainly having trouble generating a study-specific <training>.RData file. Any ideas? >>> >>> Thanks for your help. This is really cool software, and I’m excited to use it! >>> >>> Cheers, >>> David >>> >>> >>> >>> On Nov 7, 2014, at 2:23 AM, Stephen Smith <[log in to unmask]> wrote: >>> >>>> Hi - I would not necessarily recommend that you recompile the matlab - that shouldn't be necessary? >>>> >>>> The LOO option will use SGE if it's setup - so you may want to check on that - and also try running a single FIX run with a presupplied training-weights file to see if that runs OK. >>>> >>>> Cheers. >>>> >>>> >>>> >>>> On 6 Nov 2014, at 21:28, David V. Smith <[log in to unmask]> wrote: >>>> >>>>> Hi, >>>>> >>>>> I’m also having trouble setting up FIX. Our sysadmin has compiled the Matlab binaries and has changed the paths in the settings.sh file (described in the README file). >>>>> >>>>> When I train the classifier on my hand labels without the LOO, there’s no training.Rdata output file. And when I use the LOO option during training, it produces the errors below and dumps several files with strange names (e.g., "302:.e6151" and "302:.o6151”): >>>>> >>>>> leaving-out: /mnt/delgadolab/disk02/labdata02/AffInf01/FSL/203/run1.feat >>>>> /bin/sh: 1: 302:: not found >>>>> /bin/sh: 1: 302:: not found >>>>> /bin/sh: 1: 302:: not found >>>>> /bin/sh: 1: 302:: not found >>>>> /bin/sh: 1: 302:: not found >>>>> /bin/sh: 1: 302:: not found >>>>> /bin/sh: 1: 302:: not found >>>>> /bin/sh: 1: 302:: not found >>>>> >>>>> I’m not sure how to debug from here. It looks like our environment is set up correctly when I source the settings.sh file. The only other I error I see is in one of the matlab logs: >>>>> >>>>> tail: cannot open `/mnt/delgadolab/disk02/labdata02/AffInf01/FSL/203/run1.feat/fix4melview_training3_LOO_thr1.txt' for reading: No such file or directory >>>>> >>>>> But I’m assuming that file didn’t get created because something broke earlier… >>>>> >>>>> Thanks for any help! >>>>> >>>>> Cheers, >>>>> David >>>>> >>>>> >>>>> >>>>> On Nov 2, 2014, at 7:26 AM, Alireza Sedghi <[log in to unmask]> wrote: >>>>> >>>>>> Dear FSL experts >>>>>> i want to install FIX on my ubuntu 14.04 , i have downloaded it and unpacked it to my home directory , the setting.sh is somehow confusing ! i have MATLAB installed already , which part of setting.sh must change ? >>>>>> how can i change the matlab startup option >>>>>> i really appreciate if you can help >>>>>> >>>>> >>>>> ---------------------------------------------- >>>>> David V. Smith, Ph.D. >>>>> Postdoctoral Fellow, Delgado Lab >>>>> Department of Psychology >>>>> Rutgers University >>>>> Newark, NJ 07102 >>>>> ---------------------------------------------- >>>> >>>> >>>> --------------------------------------------------------------------------- >>>> Stephen M. Smith, Professor of Biomedical Engineering >>>> Associate Director, Oxford University FMRIB Centre >>>> >>>> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >>>> +44 (0) 1865 222726 (fax 222717) >>>> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >>>> --------------------------------------------------------------------------- >>>> >>>> Stop the cultural destruction of Tibet >>>> >>>> >>>> >>>> >>> >>> ---------------------------------------------- >>> David V. Smith, Ph.D. >>> Postdoctoral Fellow, Delgado Lab >>> Department of Psychology >>> Rutgers University >>> Newark, NJ 07102 >>> ---------------------------------------------- >>> >> >> >> --------------------------------------------------------------------------- >> Stephen M. Smith, Professor of Biomedical Engineering >> Associate Director, Oxford University FMRIB Centre >> >> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >> +44 (0) 1865 222726 (fax 222717) >> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >> --------------------------------------------------------------------------- >> >> Stop the cultural destruction of Tibet >> >> >> >> > --------------------------------------------------------------------------- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve --------------------------------------------------------------------------- Stop the cultural destruction of Tibet