swe_DesRep problem

[martahedl]

Hi!

I'm relatively new to SwE, but I would like to use it for my current project which is a longitudinal rs-fMRI analysis with a lot of missing data. The sandwich estimator seems like it can handle such data.

I have two groups (patients and controls). Each subject has either one or more visits (up to 5). There are 305 unique subjects and 497 inputs in total.

I've now tried to specify that model in the batch editor. I provided my 497 input files (functional connectivity maps from rs-fMRI, 3D Nifti), and additionally specified the group, visit and subject vectors as described in the documentation. I did not change any of the other default settings.

When I try to run and specify the model, it stops after just a few seconds on the error message which I'm providing below. Somehow the design matrix seems to be missing, but I don't know why SwE has problems setting that up.. What am I missing? Any help / thoughts would be much appreciated!

Thanks in advance,
Marlene

---

04-May-2022 10:14:15 - Running job #1

04-May-2022 10:14:16 - Running 'Specify Model'
Mapping files : ...done
Design configuration : ...done
Saving SwE configuration : ...SwE.mat saved
Design reporting : 04-May-2022 10:14:18 - Failed 'Specify Model'
Error using swe_DesRep (line 675)
Can't find DesMtx in this structure!
In file "/Users/Marlene/Documents/MATLAB/spm12/toolbox/SwE-toolbox-2.2.2/swe_DesRep.m" (???), function "swe_DesRep" at line 675.
In file "/Users/Marlene/Documents/MATLAB/spm12/toolbox/SwE-toolbox-2.2.2/swe_run_smodel.m" (???), function "swe_run_smodel" at line 915.

The following modules did not run:
Failed: Specify Model

[hsx1]

Hi, at first sight, it looks like you forgot to add covariates, maybe? Group, visit and subject vectors alone don't suffice. You need to create at least an intercept covariate. Either you create each one of your regressors separately under "Covariates" or you input the design matrix under "Multiple Covariates".