An epiphany lead me to the solution to the problem, which I shall post here since I am apparently due to experience it again in 2019. The NAN values in the GLM matrices were the result of there being a condition that did not appear in any of the .par files across all of the runs. Once these files were edited so that all of conditions 1, 2 and 3 appeared somewhere among these files, selxavg3-sess was able to run.
/********************************************** * Chris McNorgan * Assistant Professor * Department of Psychology * University at Buffalo, * The State University of New York * http://ccnlab.buffalo.edu/ http://ccnlab.buffalo.edu/ * Office: 716.645.0236 * Lab: 716.645.0222 **********************************************/
Hi Doug et al.,
It appears I've encountered this problem once before: https://mail.nmr.mgh.harvard.edu/pipermail//freesurfer/2016-February/043647.... https://mail.nmr.mgh.harvard.edu/pipermail//freesurfer/2016-February/043647.html
Unfortunately there is no record of any resolution (if there was one) in the original thread, and I have no recollection of what the outcome was. Where the thread did leave off was a request for the Xtmp.mat file, so I will include the Xtmp.mat file from the analysis that worked (GoodXtmp.mat) and the one that failed (BadXtmp.mat).
The error message:
Creating Design Matrix ... creation time = 0.019 sec DoMCFit = 1 ntptot = 2592, nX = 81, DOF = 2511 Saving X matrix to /home/chris/SEM/FS_0231/bold/FAM.self.sm4.down.rh/Xtmp.mat Error using svd Input to SVD must not contain NaN or Inf. Error in cond (line 35) s = svd(A); Error in fast_selxavg3 (line 279) XCond = cond(XtX);
Curiously, unless something was overlooked, the same analysis was carried out on another dataset that went through the same preprocessing steps without incident.
The Xn matrix in BadXtmp.mat contains 3 columns of NaN (:,[7,8,9]), but without knowing how those values are calculated, I can't take a principled stab at fixing the problem.
Thanks for your help, Chris
<GoodXtmp.mat><BadXtmp.mat>
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu mailto:Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.