The main problem is the 4th group where you only have two subjects and they are almost exactly the same age. This makes it hard to estimate the age slope and is causing the ill-conditioned error. You can try normalizing the ages (ie subtract the mean and divide by the stddev), you can remove age as a covariate, you can eliminate this group. Alternatively, you can elimate the age slope just for this group (this requires that you create your own design matrix).
On 8/14/18 6:00 AM, Nabin Koirala wrote:
External Email - Use Caution
Hi freesurfer community,
I am trying to do QDEC with a data for 2 groups but stuck in this error for many hours now. Tried and checked everything I could think of. Also went through the old threads and checked those recommneded but still could not solve it. Your hint will be very helpful.
/ERROR: matrix is ill-conditioned or badly scaled, condno = 18696.1/ /--------------------------------/ /Possible problem with experimental design:/ /Check for duplicate entries and/or lack of range of/ /continuous variables within a class./ /If you seek help with this problem, make sure to send:/ / 1. Your command line:/ / mri_glmfit --y /gpfs/fs1/home/mmuthura/freesurfer/subjects/qdec/Untitled/y.mgh --fsgd /gpfs/fs1/home/mmuthura/freesurfer/subjects/qdec/Untitled/qdec.fsgd dods --glmdir /gpfs/fs1/home/mmuthura/freesurfer/subjects/qdec/Untitled --surf fsaverage lh --label /gpfs/fs1/home/mmuthura/freesurfer/subjects/fsaverage/label/lh.aparc.label --C /gpfs/fs1/home/mmuthura/freesurfer/subjects/qdec/Untitled/contrasts/lh-Avg-Intercept-thickness.mat --C /gpfs/fs1/home/mmuthura/freesurfer/subjects/qdec/Untitled/contrasts/lh-Diff-f-m-Intercept-thickness.mat --C /gpfs/fs1/home/mmuthura/freesurfer/subjects/qdec/Untitled/contrasts/lh-Diff-Epi-Chr-Intercept-thickness.mat --C /gpfs/fs1/home/mmuthura/freesurfer/subjects/qdec/Untitled/contrasts/lh-X-gender-Group-Intercept-thickness.mat/ / / /2. The FSGD file (attached)/ / / /3. Design Matrix/
Design matrix ------------------ 1.000 0.000 0.000 0.000 31.300 0.000 0.000 0.000; 1.000 0.000 0.000 0.000 22.760 0.000 0.000 0.000; 1.000 0.000 0.000 0.000 29.390 0.000 0.000 0.000; 1.000 0.000 0.000 0.000 49.160 0.000 0.000 0.000; 1.000 0.000 0.000 0.000 23.810 0.000 0.000 0.000; 1.000 0.000 0.000 0.000 45.000 0.000 0.000 0.000; 0.000 1.000 0.000 0.000 0.000 21.830 0.000 0.000; 1.000 0.000 0.000 0.000 49.540 0.000 0.000 0.000; 1.000 0.000 0.000 0.000 40.550 0.000 0.000 0.000; 0.000 1.000 0.000 0.000 0.000 22.660 0.000 0.000; 1.000 0.000 0.000 0.000 26.080 0.000 0.000 0.000; 0.000 1.000 0.000 0.000 0.000 32.570 0.000 0.000; 0.000 1.000 0.000 0.000 0.000 47.190 0.000 0.000; 1.000 0.000 0.000 0.000 27.710 0.000 0.000 0.000; 1.000 0.000 0.000 0.000 25.860 0.000 0.000 0.000; 1.000 0.000 0.000 0.000 25.180 0.000 0.000 0.000; 0.000 0.000 1.000 0.000 0.000 0.000 27.630 0.000; 0.000 0.000 1.000 0.000 0.000 0.000 51.930 0.000; 0.000 0.000 1.000 0.000 0.000 0.000 25.060 0.000; 0.000 0.000 0.000 1.000 0.000 0.000 0.000 60.650; 0.000 0.000 1.000 0.000 0.000 0.000 59.360 0.000; 0.000 0.000 0.000 1.000 0.000 0.000 0.000 62.450; 0.000 0.000 1.000 0.000 0.000 0.000 20.480 0.000; 0.000 0.000 1.000 0.000 0.000 0.000 36.550 0.000; 0.000 0.000 1.000 0.000 0.000 0.000 37.000 0.000;
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer