Beth Cimini
Beth Cimini
Not literally related, but supports my 'subprocess being the culprit' theory, and search says that it's the only module code just nakedly using `subprocess` in this way. Probably wouldn't _hurt_...
Does not happen in the new dev builds. So this is a pure 4.2.6 problem.
I think a major part of the issue here is ensuring that the internals of CellProfiler recognize that disabled modules EXIST - in the GUI, they're shown, but my vague...
As we're thinking about CP5 updates to the Export modules, a reminder that this is still not ideal! [Twitter thread - requires being in BioimageAnalysis community](https://twitter.com/SBAmin/status/1590100626419351552?s=20&t=HSUX44ruR40iMBgY3Hx9Fg)
@sugan89 identified a case where this is failing, will dig in later but I now have the files to do so
This has zombies back in 4.2.4/4.2.5, in a pipeline using adaptive 3 class otsu - need to figure out why
So we've ended up in a bit of a corner case: - 3-class otsu in the skimage implementation (see linked issue above) bins the image values into a finite number...
>(the other option, of course, is to be more deterministic (I don't LOVE that actual calculation of the threshold might be different in different boxes, though in practice, since these...
Based on your internal post,  this is the line that should be barfing... https://github.com/CellProfiler/CellProfiler-plugins/blob/master/active_plugins/runcellpose.py#L429 But on a quick glance, it seems to be identical to my eye (as are...