Mihai Cara

@hbushouse > @mcara Still need to address most recent comments and fix conflict that now exists in tweakreg_step.py My apologies: somehow I thought I pushed latest changes last Friday.

> Is this second one related to this PR? Yes, see https://github.com/spacetelescope/jwst/pull/8476#discussion_r1605093709

https://plwishmaster.stsci.edu:8081/job/RT/job/JWST-Developers-Pull-Requests/1497/

'accurate' was the my original fix to the code that was originally developed by @jdavies-st (see https://github.com/spacetelescope/jwst/pull/6747). However, as @hbushouse described it in https://jira.stsci.edu/browse/JP-2612, it is "painfully slow". I wish...

@schlafly Yes, I started implementing alternative approach a little earlier. Will make a new commit soon.

Regression tests are running here: https://plwishmaster.stsci.edu:8081/job/RT/job/JWST-Developers-Pull-Requests/678/ I expect to see differences in pixel values for tests that use IVM weighting.

@schlafly @hbushouse The weighting types `'ivm-med'` and `'ivm-smed'` allow specifying window size for the median, for example: 'ivm-med5', or 'ivm-med7', ... Would that be of interest? (I thought it might,...

@hbushouse @nden What is your preference for an option to specify median filter size? see https://github.com/spacetelescope/jwst/pull/7563#issuecomment-1530095360

Why use a not-yet-released version (of numpy) when building releases?

> I was just trying to see if this would fix the problem of C-extension not on `devdeps` and it does. I know it does. I mentioned this in https://github.com/spacetelescope/spherical_geometry/pull/265#issuecomment-2106356771...