Add data source dimension name mapping option

[brynpickering]

Fixes #680

The only question I have is whether it is better described as rename_dims instead of map_dims.

Reviewer checklist

• Test(s) added to cover contribution
• Documentation updated
• Changelog updated
• Coverage maintained or improved

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 95.96%. Comparing base (b2d2dd1) to head (7589500).
Report is 16 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #687      +/-   ##
==========================================
+ Coverage   95.95%   95.96%   +0.01%     
==========================================
  Files          29       29              
  Lines        4026     4039      +13     
  Branches      845      576     -269     
==========================================
+ Hits         3863     3876      +13     
  Misses         72       72              
  Partials       91       91              

Files with missing lines	Coverage Δ
src/calliope/preprocess/data_tables.py	100.00% <100.00%> (ø)

... and 2 files with indirect coverage changes

[irm-codebase]

Looks good, but there are some CHANGELOG issues.

Additionally, a small maintainability suggestion.

[brynpickering]

@irm-codebase map_dims or rename_dims?

[irm-codebase]

@brynpickering I like map_dims a bit more.

rename_dims is a bit ambiguous (is it renaming the model's dims? the data's?). map_dims is a bit clearer, and kind of similar to some pandas functionality, so I brain likes it more ¯_ (ツ)_/¯.

[brynpickering]

@sjpfenninger I'll let you cast your vote on the naming. If it's unanimous then we'll stick with map_dims.

[brynpickering]

On the side of rename_dims, it does align with pandas rename_axis and xarray rename_dims

[irm-codebase]

Hmmm... I see the point.

Looking at how the yaml file would look like, it's perhaps less ambiguous than I thought.
rename_dims is perfectly fine!

My mind was linking the command to calliope, not pandas (timesteps is still timesteps, we are just telling data_tables to read time as timesteps), but in the yaml it is clear.

[irm-codebase]

Looking good.
Feel free to refactor to rename_dims, too!

[sjpfenninger]

I vote for rename_dims: I think it's clearer in this context