Wat ssp upd4

[adrivinca]

Required: write a single sentence that describes the changes made by this PR.

Fixes for the SSP update.
we noticed nuc_lc cooling technology was not correctly parametrized. I did not notice it because in all the scenarios I tested the technology was not active, but in some scenario it is, before 2020

Oliver suggested to pre-filter the remove.set as currently it takes lot of time just to check if the set is there

How to review

Not, sure changes are quite minimal and I already cherry picked them in ssp_dev

PR checklist

• Continuous integration checks all ✅
• whatsnew.

[codecov]

Codecov Report

Attention: Patch coverage is 93.54839% with 2 lines in your changes missing coverage. Please review.

Project coverage is 76.7%. Comparing base (44087e4) to head (3f724af).
Report is 5 commits behind head on main.

Files with missing lines	Patch %	Lines
message_ix_models/model/water/data/water_supply.py	50.0%	2 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##            main    #268     +/-   ##
=======================================
- Coverage   77.6%   76.7%   -0.9%     
=======================================
  Files        211     211             
  Lines      16079   16102     +23     
=======================================
- Hits       12481   12355    -126     
- Misses      3598    3747    +149     

Files with missing lines	Coverage Δ
message_ix_models/model/water/build.py	93.5% <100.0%> (+0.1%)	⬆️
message_ix_models/model/water/data/demands.py	78.8% <100.0%> (+0.2%)	⬆️
...essage_ix_models/model/water/data/water_for_ppl.py	95.4% <100.0%> (+0.2%)	⬆️
message_ix_models/model/water/data/water_supply.py	76.3% <50.0%> (-0.3%)	⬇️

... and 8 files with indirect coverage changes

[glatterf42]

Looks good to me, thanks :)

[glatterf42]

So this is a for inside a for inside a for? I'm surprised that this is the faster solution, it doesn't seem very efficient. Without spending more time, I'm not sure how to improve this, though, and if you say it's an improvement, then it's fine with me :)

[adrivinca]

the thing is that for any item to be removed, the message_ix_models.model.build.apply_spec() checks if elements exist and proceeds or says that there are no elements. This is very slow.
now we just exclude in advance sets from the remove list that we know are not in the scenario, so that they do not get sent to apply_spec()

[glatterf42]

If this is a serious problem and/or you know a solution for it, please record it in a new issue. This will raise awareness and eventually lead to a solution :)