non-halting states in apply_transforms

[rueberger]

when applying an affine transform using the genericLabel, apply_transforms can get stuck in a (presumably...) non-halting state pinning all cores at 100% until killed.

occurs on 0.3.2 and 0.3.3

minimal reproduction to come...

[stnava]

Yes if applied to an intensity image …. I presume?

On Tue, Jun 21, 2022 at 10:10 AM Andrew Berger @.***> wrote:

when applying an affine transform using the genericLabel, apply_transforms can get stuck in a (presumably...) non-halting state pinning all cores at 100% until killed.

occurs on 0.3.2 and 0.3.3

minimal reproduction to come...

— Reply to this email directly, view it on GitHub https://github.com/ANTsX/ANTsPy/issues/380, or unsubscribe https://github.com/notifications/unsubscribe-auth/AACPE7XLMENHAWQJJDGZWTDVQGBG3ANCNFSM5ZLSM43Q . You are receiving this because you are subscribed to this thread.Message ID: @.***>

--

brian

[rueberger]

indeed it was. is that a known issue? should be easy to catch that and throw an error in python

[gdevenyi]

It doesn't run forever. If you wait long enough, it will finish.

[rueberger]

I've run into this issue several times and it was never intentional, I think it would be helpful to throw a warning at least. Could simply throw a warning if the number of unique values in the image is beyond some value

[gdevenyi]

That would be a bug for ITK as it's an ITK filter.

[rueberger]

we can catch it in python before the ITK call