Re: [Freesurfer] glmfit - matrix is ill-conditioned or badly scaled

Linda, try demeaning your covariates and see if the problem goes away doug

On 01/30/2014 03:24 PM, Linda Zhang wrote:

Dear all,

Whilst running mri_glmfit, the following error occurred:

ERROR: matrix is ill-conditioned or badly scaled, condno = 12375.1

I've attached the design matrix, the fsgd file, and one of the contrast matrices as an example. The command line was:

mri_glmfit --glmdir lh.all.thickness.glmdir --y lh.thickness.10.mgh --fsgd all.fsgd dods --C allavovsallcvo.mtx --C avovscvo.mtx --C mavovsavo.mtx --C mavovsmcvo.mtx --C mcvovscvo.mtx --surf fsaverage lh --cortex

What should I change?

Cheers, Linda

---

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