[Simnibs-discuss] "Value Error" in charm pipeline - same issue as posted in December

[Oula Puonti]

Hi Dan,

Can you make sure that the input T1 (or scans in general) does not have
any extra singleton dimensions, e.g., shape (175,255,256,1)? If it does
just squeeze it out, write the scan back to disk and try again.

I thought we had caught all these cases, but seems that might not be
the case. We'll add this one to the list of bugs to fix.

Best,
Oula


On Thu, 2023-01-12 at 00:30 +0000, Daniel McCalley wrote:
> Hi all, 
> 
> I'm writing to report a similar instance of an error using the new
> Charm pipeline. The issue is nearly identical to the one Mariia
> Nazarova posted to the boards in December. I saw that Oula had
> followed up asking to take a peek at the T1 itself to see if there
> are any obvious issues. 
> 
> I have the same issue as Mariia (charm_log attached) and specific
> error pasted below. I reviewed my anatomical scans and did not find
> any obvious issues (screenshot below as well). 
> ValueError: could not broadcast input array from shape (175,255,256)
> into shape (175,255,256,1)
> CRITICAL: Uncaught exception
> ValueError: could not broadcast input array from shape (175,255,256)
> into shape (175,255,256,1)
> Happy to share a T1 that did and did not work through the charm
> pipeline if that would be helpful for troubleshooting! 
> 
> Best, 
> 
> Dan McCalley, Ph.D.
> Postdoctoral Scholar, Padula BRAVE lab
> Stanford University
> School of Medicine, Psychiatry and Behavioral Sciences 
> (he/him)
> _______________________________________________
> Simnibs-discuss mailing list
> Simnibs-discuss at drcmr.dk
> https://mailman.drcmr.dk/mailman/listinfo/simnibs-discuss