fix small density-bug and added test

[daniel-caichac-DHI]

Hi,

• I added an additional test for the scatter plot with density (actually extended a test by adding two lines, one with bins=19 and another with bins=21)
• The reason behind this is that sometimes, as the way the 2d-histogram for scatter density plot was defined, given certain bin-sizes, the edges of the bins (2DHist bins) ended just before the center of the bins, which crashed the plots. This could easily be solved by changing the bin-size slightly or number of bins by one, but it is not ideal.
• So now, I check the last value of the center of the bins, and the last value of the edge of the bins, and if the latter is smaller than the former, I add an extra 1/2 bin, to make sure that the bin-edges fully cover the bins-center. This has no impact on the final result (plot), but avoids the error.

[ecomodeller]

Is the double loop necessary?

How about?

hist = histodata.flatten()

[daniel-caichac-DHI]

I do not get the same result if I do that, also some tests fail. Double loop works :)
Also the histogram looks flat for some reason (no pun intended) as in, with the double loop I get this (see high density cloud in light colors):

but with the flattened histogram it just looks dark everywhere

[jsmariegaard]

Great. Thanks

Why do you use double-underscore for prefixing your helper functions? - I guess they are not that secret.

Also, would good with removing the double loop but it is not essential in my eyes

[daniel-caichac-DHI]

Great. Thanks

Why do you use double-underscore for prefixing your helper functions? - I guess they are not that secret.

Also, would good with removing the double loop but it is not essential in my eyes

Because henrik suggested making them private, I guess
#189 (comment)