I have a college who is working on some 2D data from a retinorpic map. He was attempting to adopt randomise to run on the data, which he converted to 4d NIFTI file. Unfortunately randomise does not seem to work on data in which the Z dimension is size 1.
Is there a solution to this? Here are his commands and errors:
Trying to run randomise on a 2D surface (128 x 512 pixels). 52 Subjects (31 controls, 21 patients).
nifti shape 128 x 512 x 1 x 52
print(nii.header['dim'])
array([ 4, 128, 512, 1, 52, 1, 1, 1], dtype=int16)
Command:
randomise -i output/GCL/4d.nii -o output/GCL/randomise/GCL_ -d data/glm_model.mat -t data/glm_model.con -T also randomise -i output/GCL/4d.nii -o output/GCL/randomise/GCL_ -d data/glm_model.mat -t data/glm_model.con -T2
Error:
Critical Value for: output/GCL/randomise/GCL__
tfce_corrp_tstat1 is: 0 1.91992e+14 permutations required for exhaustive test of t-test 2 Doing 5000 random permutations Starting permutation 1 (Unpermuted data)
ERROR: Program failed
An exception has been thrown
Error: tfce currently requires an input with at least 3 voxels extent into each dimension.
Colin Hawco, PhD
Neuranalysis Consulting
Neuroimaging analysis and consultation