Add a getAvailableNamedPalettes function.

[manthey]

Closes #731.

[manthey]

There are a lot of named palettes. With palettable and matplotlib, I get 3380 named palettes excluding individual colors and 4541 including them.

[banesullivan]

I get 3380 named palettes excluding individual colors and 4541 including them.

ooof. that's a lot. Too many if trying to send that over to a drop down menu in a UI

Could we separate them into a dictionary? E.g,

{
'matplotlib': ['...', ...],
'palettable': ['...', ...],
'colors': ['...', ...],
...
}

[manthey]

A large number of the variations come from (a) palettable having distinct names for each variant with a different number of points in the lookup table (e.g., cartocolors.diverging.Earth_2, cartocolors.diverging.Earth_3, cartocolors.diverging.Earth_4, cartocolors.diverging.Earth_5, cartocolors.diverging.Earth_6, cartocolors.diverging.Earth_7), and most palettes having an _r reversed variant (so cartocolors.diverging.Earth_2 and cartocolors.diverging.Earth_2_r). If we exclude the lower count palettes, the reversed palettes, and the colors, the count drops to a mere 212 palettes.

I think a better choice might be to add a reduced option to the function that would exclude the reversed palettes and the lower numbered palettes, then getAvailableNamedPalettes(includeColors=False, reduced=True) would return a much smaller set (212).

[manthey]

I pushed a change adding the reduced option. If you think this is sufficient, please approve the PR. If you still think return a dictionary of results makes sense, we probably need to discuss the granularity of it. For instance, should PIL colors and matplotlib colors be separate entries? Should we subset the palettable entries in any manner?

[manthey]

Lastly, in the reduced set of palettes, there are some function duplicates: lightbartlein.diverging.BlueGray_8 is the same as lightbartlein.diverging.BlueGrey_8; matplotlib shares 9 palettes with palettable such as Pastel2 and colorbrewer.qualitative.Pastel2_8. Short of instantiating the palettes, duplicates are somewhat hard to detect (and then you have to pick a preferred name).

[banesullivan]

I think shortening the way you did is a good implementation and this PR is good to merge.

Downstream, we may just create this list ourselves in RGD. I'm not huge on palettable and I think we'll just stick with MPL's full set (along with cmocean's as well) as the reversed maps are very useful