Hi,
One little request to greatly increase the utility of 3dABoverlap's output in many cases:
It looks like datasets A and B are each currently identified by the filename of the datasets at the time of the actual comparison within the program, which isn't necessarily the reference supplied by the user, leaving it less informative at best. At least one probable exception is that the '.gz' on compressed NIfTI file names does get included. (I say 'probable' because it's not an exception if the pipe used for the transparent decompression keeps the '.gz' in its name, though I wouldn't guess that to be the case.)
Here are two examples:
1.
Input:
3dABoverlap ../../masks_long/AAL/1006long_custom_MNI_drewVAC+orig.nii.gz 'Indices+orig[S2_Scenes-Faces_s-f_stim#0_Coef]<0..50>'
Current relevant line of output:
#A=../../masks_long/AAL/1006long_custom_MNI_drewVAC+orig.nii.gz B=./Indices+orig.BRIK
Proposed:
#A=../../masks_long/AAL/1006long_custom_MNI_drewVAC+orig.nii.gz B=./Indices+orig.BRIK[S2_Scenes-Faces_s-f_stim#0_Coef]<0..50>
(i.e. retain subscripts; the addition of '.BRIK' is good, more specific, though it would be better if that were a form that could be reused as input)
or at least:
#A=../../masks_long/AAL/1006long_custom_MNI_drewVAC+orig.nii.gz B=./Indices+orig[S2_Scenes-Faces_s-f_stim#0_Coef]<0..50>
(i.e. the literal* strings passed by the shell--no work required)
2.
Input:
3dABoverlap ../../masks_long/AAL/1006long_custom_MNI_drewVAC+orig.nii.gz '3dcalc( -a Indices+orig[S2_Scenes-Faces_s-f_stim#0_Coef] -expr ispositive(a) -datum byte )'
Current output:
#A=../../masks_long/AAL/1006long_custom_MNI_drewVAC+orig.nii.gz B=./3dcalc_PRN_QQbI2m27PY4HwxWNXRkWSQ+orig.BRIK
Proposed:
#A=../../masks_long/AAL/1006long_custom_MNI_drewVAC+orig.nii.gz B=3dcalc( -a Indices+orig[S2_Scenes-Faces_s-f_stim#0_Coef] -expr ispositive(a) -datum byte )
* One exception, if applicable: if 3dABoverlap has any internal wildcard/glob comprehension, I suspect displaying the post-expansion or resolved reference would be desirable behavior more often than not.
Thanks,
Terry
P.S. Could the '-verbose' option to 3dcalc be reserved for cases when 3dABoverlap itself is called with -verb?