AFNI version info (afni -ver): 26.1.0
I am running Surf2Surf, and compared to all other instances where I use the -prefix flag (as far as I can tell), with this program it appends the prefix to the existing dataset name. So running this:
SurfToSurf \
-i_gii std.141.${hemi}.smoothwm.gii \
-i_gii ${hemi}.smoothwm.gii \
-prefix std.141.${hemi}.${subj}.visfAtlas \
-mapfile std.141.sub-${subj}_${hemi}.niml.M2M \
-dset ${hemi}.${subj}.visfAtlas.niml.dset \
-output_params NearestNode
Gets me:
std.141.lh.03.visfAtlas.lh.03.visfAtlas.niml.dset
Instead of:
std.141.lh.03.visfAtlas.niml.dset
Specifying the file name fully with -prefix (i.e. to include niml.dset) does not address this. Am I mental, or is this different from how the -prefix flag usually works?
Howdy-
I tried a command similar to yours:
SurfToSurf \
-i_gii std.141.${hemi}.smoothwm.gii \
-i_gii ${hemi}.smoothwm.gii \
-prefix std.141.${hemi}.${subj}.visfAtlas \
-mapfile std.141.${subj}_${hemi}.niml.M2M \
-dset std.60.${hemi}.curv.niml.dset \
-output_params NearestNode
... and this produced the following two output file names:
std.141.lh.sub-000.visfAtlas.std.60.lh.curv.niml.dset
std.141.lh.sub-000.visfAtlas.1D
So, indeed, the output of the NIML dataset seems to be a combination of both the asked-for prefix and the -dset .. name. I looked in the SurfToSurf C file, it does seem to be the hardwired way to generate the output name. That is surprising to me, too.
--pt
Glad to know if I am going crazy, it's not because of this! If it is hardwired and cannot be changed, maybe it would be possible to change the documentation to warn users. In the illustrative example online (here), one would not get the impression that it has this behavior:
SurfToSurf -i rh.smoothwm.gii -i std.60.rh.smoothwm.gii \
-mapfile std.60.sub-000_rh.niml.M2M \
-dset rh.thickness.gii.dset \
-prefix my.std.60.sub-000.rh.thickness.