Hi all,
I am new to conn and am at my wits end. I keep getting this error
during the ROI extraction (step 5/7) after preprocessing: Error
using rex>rex_do mismatched number of scans and covariate
datapoints ([3 1] vs. [180 8])
I am using a custom atlas but I've checked it a million times and
everything seems correct. This happens even when I start from
scratch with just one subject. The data is structurally sound and
preprocessing completes without errors. I can verify that the
subject has 180 volumes and a TR of 2.3s, entered
correctly. I’ve tested this error with both a custom
multi-label ROI atlas (verified in FSL and SPM, 0–48 integer
labels) and with individual binary ROI masks— i'm getting the
same error in all cases. The error suggests a mismatch between the
number of scans (180) and the covariates (only 1 row?)
Any help would be greatly appreciated!
Thanks
Hi Megan,
It's not exactly clear from your description if the number 3 corresponds to the number of scans and 180 to the number of covariate samples or viceversa, if you could please copy/paste the error message that would help clarify. In any case, the issue represents a mismatch between the number of samples / timepoints that you have in your functional dataset and the number of samples / timepoints that you have in one/some of your first-level covariates. Of course the number of samples/timepoints in both of those should be the same, so I would simply figure out which one has 3 samples, as that is I imagine the incorrect one, and look in the corresponding Setup tab to identify the offending subject/session (in the Setup.functionals tab hover over the brain image and look at the description of the size of that data; and in the Setup.covariates (1st-level) tab do the same over each of the first-level covariates listed there).
Hope this helps
Alfonso
Originally posted by Megan Nakamura:
Hi all,
I am new to conn and am at my wits end. I keep getting this error during the ROI extraction (step 5/7) after preprocessing: Error using rex>rex_do mismatched number of scans and covariate datapoints ([3 1] vs. [180 8])
I am using a custom atlas but I've checked it a million times and everything seems correct. This happens even when I start from scratch with just one subject. The data is structurally sound and preprocessing completes without errors. I can verify that the subject has 180 volumes and a TR of 2.3s, entered correctly. I’ve tested this error with both a custom multi-label ROI atlas (verified in FSL and SPM, 0–48 integer labels) and with individual binary ROI masks— i'm getting the same error in all cases. The error suggests a mismatch between the number of scans (180) and the covariates (only 1 row?)
Any help would be greatly appreciated!
Thanks