Disappearing cluster alphas when Pos? box is checked

Hello AFNI peeps -

I have a stats file with a brick containing the contrast between two conditions (NOV and NUM). I can threshold the Tstat (p < .001) and cluster correct it (cluster size = 20) and when I generate the report the list of clusters will have the alpha probability values for each cluster on the right-hand side:


So far so good. But now I want to make a mask of only one side of this contrast (NOV>NUM) so I do this same process, except I also check the “Pos?” checkbox, to get rid of the NUM>NOV activations. Now when I generate the report only the positive clusters show up, as desired, but the alpha values are lost:


I can save this mask (via SaveMsk) and reload it and the mask seems to be what I want; but the absence of the alphas maked me worried that I’m doing something wrong or weird here. My 3drefit command is basically standard, aside from some scripting to allow me to automate for use with different masks and datasets:

3drefit \
-atrstring AFNI_CLUSTSIM_MASK file:${clustDir}/clustsim.${funcType}.${maskName}.mask \
-atrstring AFNI_CLUSTSIM_NN1_2sided file:${clustDir}/clustsim.${funcType}.${maskName}.NN1_1sided.niml \
-atrstring AFNI_CLUSTSIM_NN2_2sided file:${clustDir}/clustsim.${funcType}.${maskName}.NN2_1sided.niml \
-atrstring AFNI_CLUSTSIM_NN3_2sided file:${clustDir}/clustsim.${funcType}.${maskName}.NN3_1sided.niml \
-atrstring AFNI_CLUSTSIM_NN1_2sided file:${clustDir}/clustsim.${funcType}.${maskName}.NN1_2sided.niml \
-atrstring AFNI_CLUSTSIM_NN2_2sided file:${clustDir}/clustsim.${funcType}.${maskName}.NN2_2sided.niml \
-atrstring AFNI_CLUSTSIM_NN3_2sided file:${clustDir}/clustsim.${funcType}.${maskName}.NN3_2sided.niml \
-atrstring AFNI_CLUSTSIM_NN1_2sided file:${clustDir}/clustsim.${funcType}.${maskName}.NN1_bisided.niml \
-atrstring AFNI_CLUSTSIM_NN2_2sided file:${clustDir}/clustsim.${funcType}.${maskName}.NN2_bisided.niml \
-atrstring AFNI_CLUSTSIM_NN2_2sided file:${clustDir}/clustsim.${funcType}.${maskName}.NN3_bisided.niml \

and all appropriate .niml files have been created with 3dClustSim. From looking at the output of the processing pipeline, everything seems fine, no errors.

Can anyone explain where the alphas go, or if there’s a more appropriate/correct way of doing what I want to do? I’m using all AFNI tools from version 16.3.19.

Thanks for any advice,


I’ve had the same result - curious what the explanation might be?