Improve infrastructure (CI, Linting, Docs Build)

[hmgaudecker]

If there is need for speed.

[codecov]

Codecov Report

Merging #420 (19141f2) into main (79e8a0e) will decrease coverage by 0.06%.
The diff coverage is 87.55%.

@@            Coverage Diff             @@
##             main     #420      +/-   ##
==========================================
- Coverage   92.88%   92.82%   -0.07%     
==========================================
  Files         223      223              
  Lines       17396    17459      +63     
==========================================
+ Hits        16158    16206      +48     
- Misses       1238     1253      +15     

Impacted Files	Coverage Δ
src/estimagic/batch_evaluators.py	60.29% <0.00%> (ø)
src/estimagic/dashboard/callbacks.py	59.09% <ø> (ø)
src/estimagic/optimization/neldermead.py	90.62% <ø> (-0.10%)	⬇️
src/estimagic/optimization/nlopt_optimizers.py	92.66% <ø> (ø)
src/estimagic/optimization/scipy_optimizers.py	99.04% <ø> (ø)
src/estimagic/optimization/subsolvers/_trsbox.py	84.95% <ø> (ø)
...timagic/optimization/tranquilo/aggregate_models.py	87.71% <ø> (ø)
src/estimagic/parameters/check_constraints.py	76.41% <ø> (ø)
src/estimagic/parameters/kernel_transformations.py	99.20% <ø> (-0.01%)	⬇️
src/estimagic/parameters/nonlinear_constraints.py	87.93% <0.00%> (ø)
... and 73 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[hmgaudecker]

Broadening the scope a bit...

[hmgaudecker]

Some economies of scope with the review of GETTSIM #470 here -- exactly the same 😄

[janosg]

Runtime of github actions is roughly cut in half due to mamba and caching environments. I also got rid of tox entirely. @hmgaudecker maybe you want to copy this for gettsim.

[hmgaudecker]

Thanks @janosg ! Will copy, but for doing so common formatting of yaml files is really helpful. I added yamlfix and yamllint hooks, but was unable to push because @timmens was always faster than me and things interacted with his changes.

Tim, could you try to merge the yamlfix branch into this?

[janosg]

Turns out that the notebooks were not actually executed on readthedocs and shown without outputs, due to this line in conf.py:

nbsphinx_execute = "never"

Unfortunately, they also run way too long to be executed remotely. So I keep the mamba changes but otherwise revert to building notebooks locally and committing them with outputs.

[hmgaudecker]

Do all notebooks run for too long? Else use "auto" and commit only outputs of those with long runtimes?