Thanks Doug, this worked. Misconfigured my nuisance regressor seeds. Matrix still has 47 columns, but it seems to be appropriately designed.
Best, Jared
Message: 2 Date: Thu, 26 May 2016 12:14:27 -0400 From: Douglas N Greve greve@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Connectivity design ill-conditioned To: freesurfer@nmr.mgh.harvard.edu Message-ID: 574720E3.30603@nmr.mgh.harvard.edu Content-Type: text/plain; charset=windows-1252; format=flowed
It is strange that it has 47 columns, I would have expected 48, however, the current problem is caused by the wm.dat and vcsf.dat having the same content. Probably it was an error when you set up the command line for fcseed-config
On 05/25/2016 06:40 PM, Zimmerman, Jared Perry wrote:
Hi, I'm having a problem running first level connectivity analysis on my resting-state data. When I run selxavg3-sess I keep getting a notice that my design is ill-conditioned but I can't figure out why. I am trying to run analysis on two 124 volume rsfMRI runs in my fsd and the design matrix looks fine to me with dimensions 248x47 (although should ncols be an even number since I'm doing two runs?).
Poking around the Xtmp.mat I noticed a few things that are confusing. Notably, flac.funcfspec points only to one of my runs in the fsd, and flac.X is a 124x24 matrix which appears to be the design matrix for just one of my runs (and if there are 24 columns in that, should there be 28 columns in the design matrix for combining both runs?). Any help figuring out why this design is ill-conditioned would be great.
I've attached my Xtmp.mat, selxavg3.log and analysis design file to this email, but it's unclear to me if the mailing list will scrub them off. If so I can make them available some other way.
Thanks, Jared
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer