setBatchImageCalculation crashes during RFX

[CerenB]

Describe your problem in detail

I am running
bidsRFX('RFX', opt, funcFWHM, conFWHM);

and when it comes to setBatchImage step, it crashes.

Error message

BUILDING JOB: image calculation
Error using setBatchImageCalculation (line 62)
The value of 'matlabbatch' is invalid. It must satisfy the function: iscell.

Error in setBatchMeanAnatAndMask (line 67)
  matlabbatch = setBatchImageCalculation(matlabbatch, ...

Error in bidsRFX (line 69)
      matlabbatch = setBatchMeanAnatAndMask(matlabbatch, ...
 

When I check in line#62

62    parse(p, varargin{:});
K>> parse(p, varargin{:})
The value of 'matlabbatch' is invalid. It must satisfy the function: iscell.

apparently,
p.Results is a struct with no fields.

Any ideas?

[Remi-Gau]

I think that initializing the batch with an empty cell here should do.

https://github.com/cpp-lln-lab/CPP_SPM/blob/a9d8c618c5d30d83a148c864f602268f211587a3/src/workflows/bidsRFX.m#L68

matlabbatch = {};

Can you try and let me know?

I am slowly trying to start using argparse more in the code base but I have not done it everywhere yet.

[Remi-Gau]

Ideally we want to open a PR to change all the matlabbatch = []; :-)

[CerenB]

oh, OK I was trying to go back in the branches to see what happened! it works before in this modification.

Now I tried with your suggestion, it worked! thank you so much for quick feedback!

[Remi-Gau]

I pushed a commit to fix this on the big fMRIprep PR.

bf61773

[CerenB]

ok, shall I close this issue?

in fmriprep_input branch, bids-matlab submodule etc has to be updated compared to dev branch right? is it not recommended to work in fmriprep branch yet?

[Remi-Gau]

yeah the fmriprep branch is not quite yet done.

should be soon, but merging it should trigger a new version release because it will change a lot of things.

[Remi-Gau]

I can open a PR on the dev branch and just cherry pick the commit I applied on the fmriprep branch.

[CerenB]

I did my own branch to keep this change, so if it's only me, then please don't bother. (I'm slightly worried of submodule crashes :))

[Remi-Gau]

na it's fine, your branch is based on dev.
I can make a quick and easy PR.

[CerenB]

I did try on my local, everything works!