Add a feature to pass an explicit weight dataframe to downscale_region()

[danielhuppmann]

Please confirm that this PR has done the following:

• Tests Added
• Documentation Added
• Description in RELEASE_NOTES.md Added

Description of PR

This PR adds an option to pass an explicit weight-dataframe with region (and possibly others) as index dimension and the time domain as columns. Implementation based on many discussions with @maartenbrinkering on his specific use case.

[codecov]

Codecov Report

Merging #411 into master will increase coverage by 0.06%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #411      +/-   ##
==========================================
+ Coverage   93.26%   93.33%   +0.06%     
==========================================
  Files          35       35              
  Lines        4129     4169      +40     
==========================================
+ Hits         3851     3891      +40     
  Misses        278      278              

Impacted Files	Coverage Δ
pyam/core.py	92.24% <100.00%> (+0.10%)	⬆️
tests/test_feature_downscale.py	100.00% <100.00%> (ø)

---

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 911dcd4...1f34641. Read the comment docs.

[danielhuppmann]

@fabiosferra, maybe also of interest to you?

[maartenbrinkerink]

Approved. Based on the tests we've done the feature works as intended.

[l-welder]

Tested the new features locally and everything seems to be working as intended. Looking forward to add to this (=

[l-welder]

looks great!

on the risk of appearing meticulous: passing the default values of optional arguments is not 100% consistent in the doc string (e.g. region lost "World")... but I see that is also the case for other docstrings and not really of importance.... but I had to find something (=

[danielhuppmann]

Thanks @l-welder for the review and noticing the change in the docs. We actually started out specifying the default in the docstrings, but (if I recall correctly) @khaeru pointed out that the NumPy-style docstrings convention is to just write "optional", because the default value is anyway obvious from the function signature. So I'm changing this whenever I stumble over the legacy docstrings...