New grayscale style

[aloctavodia]

Adds new grayscale style. Originally intended for the BMCP book but probably useful for a more general use. It is based on the arviz-white style, but the color cycle has only five colors the first four are grays and the fifth is the ArviZ's "C0 blue", the cmap is matplotlib's gray cmap

• Follows official PR format
• Includes a sample plot to visually illustrate the changes (only for plot-related functions)
• New features are properly documented (with an example if appropriate)?
• Code style correct (follows pylint and black guidelines)
• Changes are listed in changelog

A few examples from the gallery (some of with their default colors changed)

[codecov]

Codecov Report

Merging #1164 into master will increase coverage by 0.28%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #1164      +/-   ##
==========================================
+ Coverage   93.11%   93.39%   +0.28%     
==========================================
  Files          94       94              
  Lines        9289     9995     +706     
==========================================
+ Hits         8649     9335     +686     
- Misses        640      660      +20     

Impacted Files	Coverage Δ
arviz/plots/violinplot.py	90.74% <0.00%> (-0.17%)	⬇️
arviz/plots/forestplot.py	94.00% <0.00%> (-0.12%)	⬇️
arviz/plots/essplot.py	100.00% <0.00%> (ø)
arviz/plots/mcseplot.py	100.00% <0.00%> (ø)
arviz/plots/autocorrplot.py	100.00% <0.00%> (ø)
arviz/utils.py	91.18% <0.00%> (+0.15%)	⬆️
arviz/stats/stats.py	96.96% <0.00%> (+0.63%)	⬆️
arviz/plots/traceplot.py	98.19% <0.00%> (+0.86%)	⬆️
arviz/plots/rankplot.py	92.85% <0.00%> (+0.96%)	⬆️
arviz/plots/ppcplot.py	96.63% <0.00%> (+1.34%)	⬆️
... and 4 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 6a51574...bbf2d37. Read the comment docs.

[ahartikainen]

what cmap is this? The mpl default or the one in colorcet (etc+)

[aloctavodia]

The default from matplotlib, as I understand this one is perceptually uniform. But I could check again.

[ahartikainen]

It is not.

[aloctavodia]

:( OK, I will fix it

[aloctavodia]

updated cmap figures

I was unsure if we should use the linear_grey_10_95_c0 or linear_grey_0_100_c0 scales. The first one does not include the extreme values, i.e the white and black. @ahartikainen What do you think?

[canyon289]

This is nice!

[canyon289]

updated cmap figures

I was unsure if we should use the linear_grey_10_95_c0 or linear_grey_0_100_c0 scales. The first one does not include the extreme values, i.e the white and black. @ahartikainen What do you think?

If this is meant for printed book format should we ask publishers which plot will look better in print? My assumption for book is if readers decide to reproduce on computer the limitation of grayscale wont apply anymore and we tell them to use full colorscale, so this styles main use case is printed material

[ahartikainen]

Also, for printing, we can force colours to be a bit lighter

[aloctavodia]

This is more general than the book. Many journals use gray-scale figures or charge for color ones. So having an easy to use grey-scale could be useful for many users. Notice this adds the cet_grey_r and cet_grey_r` cmaps.