Sorry, I thought I sent an email responding to this. Try running it with -run-wise. This will analyze each run separately. I can't remember whether it generates a final combined average (been a long time since I used this option), but run it and see what it does.

On 6/23/2022 2:00 PM, Caspar M. Schwiedrzik wrote:

External Email - Use Caution

Hi!

Any advice on this issue?

The error message is:

Starting contrasts
UrInv J=1 -------------
Out of memory. Type "help memory" for your options.

Error in fast_glm_pcc (line 39)
yhatd = Rd*(X*beta);

Error in fast_fratiow (line 136)
if(nargout == 6) pccbin = fast_glm_pcc(beta(:,indbin),Xbin,C,rvarbin);
end

Error in fast_selxavg3b (line 1148)
fast_fratiow(betamat,X,rvarmat,C,acfsegmn,acfseg.vol(:));

>> ------------------------------------------
ERROR: fast_selxavg3() failed\n

Thank you,

Caspar

Am Mo., 20. Juni 2022 um 10:17 Uhr schrieb Caspar M. Schwiedrzik <cschwiedrz@rockefeller.edu>:

Hi!

I am running large single subject GLMs with selxavg3-sess and I run into a Matlab "out of memory" error. I was wondering whether you had any advice on how to get around that.

I have 40-50 runs per subject, and the analyses are whole brain.

The GLM per subject runs through fine, but the error occurs once selxavg3-sess starts computing contrasts.

The problem occurs with v5.3 and v7.2.

This is a Centos 7 system with 320 GB of memory, and 7 GB of swap memory.

Any advice would be appreciated!

Caspar
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer