ENH: Follow ups for the hilbert_distance function about dask-geopandas

[martinfleis]

@TomAugspurger We don't depend on it. It is used only in CI. We use the custom numba implementation ported from spatialpandas.

[martinfleis]

One more thing to add is to ensure that codecov sees the coverage of numba jitted functions. We'll probably need to run that bit again without that using env variable.

[tastatham]

Another follow up is to centralise/ generalise several functions that are used by both hilbert_distanceandmorton_distance`.

• Firstly, the _continuous_to_discrete_coords function is imported from hilbert_distance.py in morton_distance.py - centralise
• Secondly, the _distances_from_coordinates function is used in both hilbert_distance.py in morton_distance.py but are defined in slightly different ways - generalise

[TomAugspurger]

I didn't follow this implementation closely, but is the plan to rely on the hilbertcurve package long-term? If so, it needs to be added to the install_requires:

dask-geopandas/setup.py

Lines 6 to 11 in 99fa1a8

	install_requires = [
	"geopandas",
	"dask>=2.18.0,!=2021.05.1",
	"distributed>=2.18.0,!=2021.05.1",
	"numba",
	]

.