Traceplot legend

[OriolAbril]

Description

Fixes #837. Allows customization of properties to distinguish lines in the same subplot with all combinations of combine and compact. Introduces arguments compact_prop and chain_prop with format (property_name, (value1, value2...)) to set color, linestyle, linewidth...

Matplotlib version is kind of working, only labels are missing, plot format is OK. Bokeh is still more preliminar. It would also be interesting to implement interactive legends in bokeh for traceplot, but I currently do not plan to do this here.

Checklist

• Follows official PR format
• New features are properly documented (with an example if appropriate)?
• Includes new or updated tests to cover the new feature
• Code style correct (follows pylint and black guidelines)
• Changes are listed in changelog

[OriolAbril]

Current behaviour:

1. combined=False, compact=False (default)

2. combined=True, compact=False

3. combined=False, compact=True

4. combined=True, compact=True

How does this look? cc @canyon289 @julian-pani

[codecov-io]

Codecov Report

Merging #1070 into master will decrease coverage by 2.25%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #1070      +/-   ##
==========================================
- Coverage   93.50%   91.24%   -2.26%     
==========================================
  Files          93       93              
  Lines        8773     8900     +127     
==========================================
- Hits         8203     8121      -82     
- Misses        570      779     +209     

Impacted Files	Coverage Δ
arviz/data/io_pystan.py	56.14% <0.00%> (-42.88%)	⬇️
arviz/plots/backends/__init__.py	29.11% <0.00%> (-20.26%)	⬇️
arviz/data/io_emcee.py	93.33% <0.00%> (-5.00%)	⬇️
arviz/plots/backends/bokeh/distplot.py	84.31% <0.00%> (-3.93%)	⬇️
arviz/data/io_cmdstanpy.py	100.00% <0.00%> (ø)	⬆️
arviz/plots/backends/matplotlib/forestplot.py	96.53% <0.00%> (+0.01%)	⬆️
arviz/plots/plot_utils.py	94.82% <0.00%> (+0.08%)	⬆️
arviz/plots/backends/bokeh/forestplot.py	92.88% <0.00%> (+0.02%)	⬆️
arviz/stats/stats_utils.py	91.46% <0.00%> (+0.02%)	⬆️
arviz/data/inference_data.py	83.13% <0.00%> (+0.06%)	⬆️
... and 6 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 d57b851...408d8fb. Read the comment docs.