Can you check the modification time of ~/audio/all/audio.lh/tones/ces.nii.gz ? Try deleting it and rerunning isxconcat-sess. The only other thing that I can think of is that some of the individual ces files have multiple frames. You can check that with mri_info --nframes */bold/audio.lh/tones/ces.nii.gz If the problem persists, send the terminal output of isxconcat-sess

On 4/5/19 8:05 AM, Keri Woods wrote:

External Email - Use Caution

Hi Freesurfer experts

I'm having a problem running an FSFAST group analysis. When I try to run the second level GLM, I get the following error:

keri@buddha:~/audio_108_no_stc_tau1,5$ mri_glmfit --y ~/audio/all/audio.lh/tones/ces.nii.gz   --wls ~/audio/all/audio.lh/tones/cesvar.nii.gz   --fsgd 2groups.fsgd  --C groupdiff.mtx   --surface fsaverage lh   --glmdir all_tones_left_glm.wls   --nii.gz gdfRead(): reading 2groups.fsgd INFO: gd2mtx_method is dods Reading source surface /home/keri/recon_all/subjects/fsaverage/surf/lh.white Number of vertices 163842 Number of faces    327680 Total area         65416.984375 AvgVtxArea       0.399269 AvgVtxDist       0.721953 StdVtxDist       0.195470

$Id: mri_glmfit.c,v 1.241.2.4 2016/12/08 22:02:40 zkaufman Exp $ cwd /home/keri/audio_108_no_stc_tau1,5 cmdline mri_glmfit.bin --y /home/keri/audio/all/audio.lh/tones/ces.nii.gz --wls /home/keri/audio/all/audio.lh/tones/cesvar.nii.gz --fsgd 2groups.fsgd --C groupdiff.mtx --surface fsaverage lh --glmdir all_tones_left_glm.wls --nii.gz sysname  Linux hostname buddha machine  x86_64 user     keri FixVertexAreaFlag = 1 UseMaskWithSmoothing     1 OneSampleGroupMean 0 y    /home/keri/audio/all/audio.lh/tones/ces.nii.gz logyflag 0 usedti  0 FSGD 2groups.fsgd labelmask /home/keri/recon_all/subjects/fsaverage/label/lh.cortex.label maskinv 0 glmdir all_tones_left_glm.wls IllCondOK 0 ReScaleX 1 DoFFx 0 wFile /home/keri/audio/all/audio.lh/tones/cesvar.nii.gz weightinv  1 weightsqrt 1 Creating output directory all_tones_left_glm.wls Loading y from /home/keri/audio/all/audio.lh/tones/ces.nii.gz    ... done reading. INFO: gd2mtx_method is dods Saving design matrix to all_tones_left_glm.wls/Xg.dat Computing normalized matrix Normalized matrix condition is 1 Matrix condition is 1.34783 Found 149955 points in label. Pruning voxels by thr: 1.175494e-38 Found 66097 voxels in mask Saving mask to all_tones_left_glm.wls/mask.nii.gz Reshaping mriglm->mask... search space = 31233.070725 ERROR: dimension mismatch between y and X.   y has 116 inputs, X has 108 rows.

108 is the number of subjects that I have. I can see that the y file with 116 inputs is /home/keri/audio/all/audio.lh/tones/ces.nii.gz , and checking the dimension of it, it confirms that it is a 4D matrix with the 4th dimension of 116. It seems as if this file is created during isxoconcat-sess, and I have checked and the session ID file being used here has only 108 entries. It appears that 108 subjects are being used in this step, as part of the output is:

ninputs = 108 Checking inputs nframestot = 108 Allocing output Done allocing nframes = 108

I think that a source of the error could be the fact that at one time I did have 116 subjects. I've created a new folder, copying in the unpacked files for each subject, and repeating from the preprocessing step, but this doesn't make any difference. Is there are chance that there's a file outside the functional directory that has stored this 116?

Any help is greatly appreciated.

Best wishes, Keri

---

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