(bug) support graph in moran's viz

[weikang9009]

The PR is to fix the issue in calculating spatial lag when the Moran instance has the graph weight type

[martinfleis]

Fine with this change but as I believe @ljwolf pointed out, we can do this fix over in libpysal lag_spatial ensuring that it produces correct lag irrespective of the weights type. I'll open a PR there as well.

I am a bit hesitant to fix stuff in splot given it is schedules for disassembling and sunsetting.

[weikang9009]

Fine with this change but as I believe @ljwolf pointed out, we can do this fix over in libpysal lag_spatial ensuring that it produces correct lag irrespective of the weights type. I'll open a PR there as well.

Sounds like a much better idea!

I am a bit hesitant to fix stuff in splot given it is schedules for disassembling and sunsetting.

Will the viz functions be moved to the other pysal packages?

[martinfleis]

Will the viz functions be moved to the other pysal packages?

The idea is to move esda viz to esda, giddy viz to giddy etc, to ensure it is maintained alongside the classes it attempts to visualise. splot has been a bit abandoned...

[knaaptime]

funny timing, i was just hitting this myself and resorting to to_W because I didn't want to open an splot pr :P

[weikang9009]

resolved via pysal/libpysal#765