Refresh convert.py

[eantonini]

Closes #362 .

Changes proposed in this Pull Request

This PR implements a refreshed convert.py with hopefully clearer arguments and logic. Tests and Jupyter notebooks have been updated to reflect these changes.

• Newly introduced arguments are:

  • mean_over_time in place of capacity_factor
  • sum_over_time
  • capacity_units

• Arguments that change meaning are:

  • capacity_factor in place of per_unit

• Removed arguments are:

  • capacity_factor_timeseries
  • per_unit

• A new logic has also been introduced to account for cases where aggregation is to be performed (aggregate=True when matrix, shapes or layout is passed) and cases where the capacity and capacity factor have to be computed after the aggregation (capacity_factor=True or return_capacity=True).

• convert_temperature, convert_soil_temperature, and convert_dewpoint_temperature have been combined into a single convert_temperature because all those functions were simply converting Kelvin to Celsius. They don't seem to be used outside convert.py.

• Lastly, units are now specified in each of the convert_* function.

Checklist

• Code changes are sufficiently documented; i.e. new functions contain docstrings and further explanations may be given in doc.
• Unit tests for new features were added (if applicable).
• Newly introduced dependencies are added to environment.yaml, environment_docs.yaml and setup.py (if applicable).
• A note for the release notes doc/release_notes.rst of the upcoming release is included.
• I consent to the release of this PR's code under the MIT license.

[euronion]

I'll try to have a look next week. Some brief comments:

• Are there overlaps with / does it supersede Fix capacity factor timeseries in convert.py #346 ?
• We'll have to look carefully at the backwards compatability / breaking changes
• Type hints to the changed code would be nice, even without static typing (related Static typing #472 Type hints, docstrings, comments #471 )
• We'll need good release notes
• New tests for sum_over_time and capacity_units
• Adjustments to documentation and release notes as soon as we are ready.

[eantonini]

Hey @euronion, thanks for your feedback and suggestions.

Since your last comment, I added type hints and checked any overlaps with #346 . Both PRs implement a clearer logic for aggregation. From that PR, I copied the case in which a default capacity layout (1MW per grid cell) is used if no layout or matrix is provided. That PR also adds a feature to resample the output time series. Let me know if that is a feature that you also want to be implemented.

All tests now run error free.

[FabianHofmann]

thanks for the PR @eantonini and sorry things take so long here. Such a refactor on the UX is definitely needed. I am wondering if we can make things more structured and transparent if another set of arguments, such as aggregate_time: Literal["sum", "mean", bool], aggregate_space = bool. I will come back to this after I have spent more thoughts on this. But that said it might be that this approach proposed here won't make it (just as a heads up)

[FabianHofmann]

I am closing this as we should go for a re-stucturing like this #489