shnizzedy
commented
2 years ago Links to the referenced lines above
When the cwas.py script loads the 4-D functional images into pandas, it reads each subject into an array of (subject, volumes (time), x, y, z) (https://github.com/FCP-INDI/C-PAC/blob/55ae58c004307294dae41ba4a60120ca444bf465/CPAC/cwas/cwas.py#L166-L169) based on a mask of the subject group (if you allow it auto-create a mask or input an ROI mask).
In the function
calc_subdists(https://github.com/FCP-INDI/C-PAC/blob/55ae58c004307294dae41ba4a60120ca444bf465/CPAC/cwas/cwas.py#L48-L61) gives this error"ValueError: not enough values to unpack (expected 3, got 1)".
Describe the bug
Hello! I have been looking at the code for running MDMR through cpac's group analysis feature. I however ran into a roadblock that I can't seem to understand or at least figure out. When the cwas.py script loads the 4-D functional images into pandas, it reads each subject into an array of (subject, volumes (time), x, y, z) (line 166; cwas.py) based on a mask of the subject group (if you allow it auto-create a mask or input an ROI mask). This becomes a problem because in the code for computing the distance matrix there is an error on how the voxel information is extracted. In the function calcsubdists (line 48; cwas.py) subjects, voxels, = subjects_data.shape gives this error "ValueError: not enough values to unpack (expected 3, got 1)". The output has (37, ) which is missing the voxel information. Somewhere it deletes it or loses the information. (37 is the number of subjects I have running even though I asked for 33).
To reproduce
Run the command cpac group mdmr [group_config.yml] and let the code run. Input 4D nifti files from CPAC 1.7.2 functional_to_standard_smooth folder. This has to be run under CPAC 1.7.2 to avoid the yaml parsing error that appears.
Expected behavior
For the matrix calculation to occur like the code runs in the R counter part. Then to calculate the correlation and output a z-score image of the voxels associated with the variable of interest accounting for the selected regressors.
Acceptance criteria
Testing the port from R to python to make sure no other bugs appear.
Screenshots
No response
C-PAC version
v1.7.2
Container platform
Docker
Docker and/or Singularity version(s)
No response
Additional context
No response