It sounds like there's something wrong with your design matrix. Try loading it into matlab and looking X'*X. If there are significant off-diagonal values, then that means that those components are too correlated to estimate.
doug
Goldman, Aaron (NIH/NIMH) [C] wrote:
Thanks. I had one more question now that I'm to mri_glmfit. Based on an error message I'm getting, it seems something about my design matrix isn't set up right. I'm inputting a text file for x in which each row looks like:
1 0 0 1 0 42.35
with the first three numbers being my class variable of interest, the next two gender, and the last age. It tells me the "matrix is ill-conditioned or badly scaled, condno = 1e+08," and then prints out the matrix and aborts. Do you know what I need to do differently for this to work? Thanks,
-Aaron-
-----Original Message----- From: Doug Greve [mailto:greve@nmr.mgh.harvard.edu] Sent: Wednesday, February 14, 2007 11:49 AM To: Goldman, Aaron (NIH/NIMH) [C] Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] smoothing question
something like:
mri_surf2surf --s subject --hemi lh --sval input --tval output --fwhm 5
doug
Goldman, Aaron (NIH/NIMH) [C] wrote:
Hi, I'm attempting to do thickness analysis based on the guide at:
http://surfer.nmr.mgh.harvard.edu/docs/ftp/pub/docs/fsgroupana.pdf
For the smoothing part using mri_surf2surf (page 14), I was wondering what a typical command line should look like.
Thanks,
-Aaron-
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer