Error with palm_gcdf

[kellyhiersche]

I am trying to do a group analysis using PALM using the .dtseries.nii provided in the HCP S1200 release. I am new to processing fMRI data, so I am trying to do a test run by only analyzing one lower level COPE image from 5 subjects. I have been trying to follow the HCP TaskfMRI Analysis tutorial: https://wustl.app.box.com/v/hcp-2019-practical-11, but the practice data was not available for download, so I attempted to create the relevant files for my data.

I am running the code:

palm -i Cope10.dtseries.nii -transposedata -d designmatrix.mat -t C.con -o results -logp

And getting

-i Cope10.dtseries.nii

-transposedata

-d designmatrix.mat

-t C.con

-o results

-logp

Found FSL in /usr/local/fsl

Found HCP Workbench executable in /Applications/workbench/bin_macosx64/wb_command

Reading input 1/1: Cope10.dtseries.nii

Reading design matrix and contrasts.

Elapsed time parsing inputs: ~ 1.47814 seconds.

Number of possible permutations is 120 = 5!.

Generating 120 shufflings (permutations only).

Building null distribution.

0.208% [Design 1/1, Contrast 1/4, Shuffling 1/120, Modality 1/1]

Saving file: results_dat_tstat_c1

Error using palm_gcdf (/Applications/palm-alpha116/palm_gcdf.m:63)

X must be in the interval [0,1].

Error in palm_gtoz (/Applications/palm-alpha116/palm_gtoz.m:80->palm_gcdf)

Error in palm_core (/Applications/palm-alpha116/palm_core.m:1599->palm_gtoz)

Error in palm (/Applications/palm-alpha116/palm.m:81->palm_core)

I tried looking at both the palm_core.m and palm_gcdf.m, but could not figure out why I am getting that error, or how to fix it.
I am pretty sure that my design matrix, dt.series.nii, and contrast files are all properly set up.

Thoughts? Has anyone run into this error?
*I tried to post this to the FSL listserv, but I don't not have access to post questions. I did search prior posts, but did not see anything with the same error.

[andersonwinkler]

Hi Kelly,

Sorry the delay. Every time this issue appeared, it was at the end a problem with the data... However, just to be sure, are you using Matlab or Octave? And which version?

If Octave, try to use at least 4.2.0 if you could.

Thanks!

Anderson