Hello Doug,
Thanks so much for fixing this bug. I tried it and it works perfectly fine and it solves the problem that I had.
Many thanks again, Nasim
Nasim and Judith (and
everyone else),
I've found and fixed a bug that was
causing the surfcluster results for
the mc and perm simulations
of absolute "abs" contrasts.
To implement
the fix, get Linux 64bit versions of these files and copy
them
to $FREESURFER_HOME/bin (make backups of the originals:).
ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/mri_surfcluster
ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/mri_glmfit
You can then run something like:
mri_glmfit-sim --glmdir yourglmdir --no-sim mc-z.abs
The "--no-sim" option tells it NOT to do the full simulation again, just
use the CSDs that are there to recompute the cluster
results for each
contrast, so it should be done in a matter of
minutes.
If you've used a permutation, then run
mri_glmfit-sim --glmdir yourglmdir --no-sim perm.abs
Please let me know if this fixes the problem.
doug