Open jGaboardi opened 6 years ago
I'm into this, but I don't have time to do this myself. Right now, what I plan to do going forward is to continue subclassing scikit-style estimators in clustering, adopting their API as much as possible, and hoping that in future this consistency ensures that everything can be repackaged together with similar behavior.
It's working internally on my research code base right now, so I think we may want to consider this for regionalization, too. I can share examples, but spenc/skater do this.
If we do go forward with a global spatial optimization submodule I would be able to contribute to the facility location modeling side of things, but I don't have the time or expertise to do the any of the regionalization stuff.
That's fine. the point of keeping everything together would be to ensure code reuse. I only recently have been working with heuristics here, and I'm not sure there's reuse between the heuristics and the existing MIP stuff here.
Maybe we should chat about this with other devs at a call next time.
Updating that we are going to end of life region with an upcoming release and point users to pysal/spopt from there.
Explore potential workload / heavy lift for creating a new pysal submodule for spatial optimization (
pysal/spopt
?) and merging inpysal/region
as a component.Keep this in mind for the near future.
@ljwolf @yogabonito