I've been thinking about this a bit. I am inclined to think that the level sets stuff should move to data_objects like you suggest, and for now, not change the API. But, moving forward, I think we *should* change it a bit (but not much) to flatten the object hierarchy a tad, and make the clump objects match the extracted sets more -- augment the two to make one work for both. For now, just moving (with a notice about the move) should be fine.
-Matt On Fri, Aug 3, 2018 at 4:48 PM Britton Smith email@example.com wrote:
Today we accepted a pull request backporting all changes from yt's analysis_modules to their future home in the yt_astro_analysis package. This makes yt_astro_analysis ready for a stable release from a codebase perspective, although there are still a few more things to do. The current plan (unless their are other opinions) is to announce a stable release of yt_astro_analysis at the same time as the next release of yt. This also makes sense since yt_astro_analysis currently requires the development version of yt.
This leaves the level_sets module (countouring and clump finding) as the last non-deprecated package in the analysis_modules directory. A discussion from a long time ago concluded with the idea that level_sets be moved to somewhere in yt/data_objects. What was not decided is where in data_objects, whether it would take on a new name, and if the API should change at all.
On Slack, Nathan suggested having something like ds.clump_tree. Does anyone else have any thoughts?
Britton _______________________________________________ yt-dev mailing list -- firstname.lastname@example.org To unsubscribe send an email to email@example.com