On 4/6/2020 10:57 AM, s7675854@stud.uni-frankfurt.de wrote:
External Email - Use Caution Hi Douglas,
thanks a lot for your quick response. I wanted to use Monte Carlo sim as all other analyses in our project include a monte carlo simulation and I try to keep things consistent. Am I getting you right - there is no way to run a monte carlo simulation with LGI data? And why exactly would you not recommend it in this case (even if it worked)?
Yes, you can do the MCZ simulation yourself with mri_mcsim, something like mri_mcsim --o /your/space/mcsim/fsaverage/lh/cortex --base mc-z --save-iter --surf fsaverage lh --nreps 1000 --fwhm 37 Then run mri_glmfit-sim as you did before, but specify --mczsim-dir /your/space/mcsim
I did try permutations with freesurfer in the past, but it never really worked out for me, to be honest. They kept running for ages and I stopped them after 6-7 days...
Not sure what happened, but if you want to debug it, I can help
Thanks in advance,
Doro Zitat von "Douglas N. Greve" dgreve@mgh.harvard.edu:
The problem is that mri_glmfit-sim is looking for a entry in the table that does not exist. The table only goes up to a smoothness level of 35. LGI data tend to bed very smooth. I would strongly recommend that you use permutation instead. See https://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/MultipleComparisonsV6.0...
On 4/6/2020 7:42 AM, s7675854@stud.uni-frankfurt.de wrote:
External Email - Use Caution Dear Freesurfer experts,
I'm trying to run a simulation (--> mri_glmfit-sim) with cached LGI data (fwhm10), but get an error due to the estimated fwhm kernel. The folder corresponding to the fwhm value (-->37) does not exist, I guess because it's a pretty unusual value...?
DoSim = 0 UseCache = 1 DoPoll = 0 DoPBSubmit = 0 DoBackground = 0 DiagCluster = 0 gd2mtx = dods fwhm = 36.855100 ERROR: cannot find /apps/prod/x86_64/freesurfer-6.0.1/average/mult-comp-cor/fsaverage/lh/cortex/fwhm37/abs/th20/mc-z.csd
I've seen mails describing similar issues in the archive and from your explanation I more or less understand that you cannot define this value as it is estimated in the process...? However, I did not see if there is a solution to this problem. What could I do about it now?
I attached the logfiles and would be very grateful for your help,
thanks a lot and stay healthy,
Doro
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer