Dear FSL Experts,
To follow up on this prior thread, we still find ourselves trying to run our very large ANOVA with more than 2000 copes. I have two questions where any advice would be helpful:
1) Fslmerge. I believe I've created a .fsf file that is complete, and feat_model runs appropriately. However, the process seems to fail on fslmerge. As possible, I have run the merges ahead of time (on the mask, mean_func, cope 1, varcope1, etc) and placed them in the appropriate path. However, I am limited by the fact that the fslmerge commands seem to be missing from the log file. Could you provide any insight into why these commands are not in the log?
2. Filtered_func: the filtered_func_data file is missing, but it's not readily apparent to me from looking at smaller, completed ANOVA logfiles where and how this file is
created.
I am limited in debugging by the fact that it takes somewhere around 180 hours of processing time on our best machine (32 GB memory) to create the .mat file and then begin the anova (and then fail). Thus, any insight into how to fix this problem before continuing trial-and-error troubleshooting would be exceptionally helpful.
I would very much appreciate any help or insight you can provide.
Regards,
Mark Griffin, BA
Brain Behavioral Laboratory
Department of Psychiatry
University of Pennsylvania School of Medicine
|