Reorganize examples

[benegee]

Based on #29, I suggest the following structure for a start.

level 1

distinguish between equations

Current choice:

• advection
• euler
• shallowwater

level 2

distinguish based on geometrical concept

Current choice:

• cartesian
• covariant

(for the time being there is only cartesian under euler, which therefore left out)

euler/cartesian level 3

distinguish based on supported species

Current choice:

• dry_air
• moist_air
• precipitation

cartesian_euler level 4

distinguish between test cases

Current choice:

• buoyancy
• global_circulation
• mountain_flow
• tests

Any feedback highly appreciated, we can still change everything!

[codecov]

Codecov Report

❌ Patch coverage is 89.55224% with 7 lines in your changes missing coverage. Please review.
✅ Project coverage is 97.10%. Comparing base (bb81c47) to head (5f2add1).
⚠️ Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
src/equations/compressible_moist_euler_2d.jl	89.55%	7 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #135      +/-   ##
==========================================
- Coverage   97.21%   97.10%   -0.12%     
==========================================
  Files          33       33              
  Lines        4491     4558      +67     
==========================================
+ Hits         4366     4426      +60     
- Misses        125      132       +7     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[benegee]

I also

• introduced corresponding sub-testsets to get a more detailed test summary output
• removed some moist Euler tests, which very were similar to others and not needed for coverage
• moved initial_condition_moist_bubble, which appeared in two elixirs, to equations/ and used this in both elixirs (this causes coverage to decrease because the initial condition has some safeguards, which do not get triggered, and this now affects coverage because I moved it to src/)

[tristanmontoya]

Nice work @benegee! What are your thoughts on just having advection, shallow_water, and euler as the top level (e.g. what physics are being solved)?

For shallow water and maybe advection (although less necessary there because there are fewer tests) there could be a second level of cartesian and covariant. Once we have a covariant Euler solver, there could be a second-level division there too.

I feel like users of TrixiAtmo.jl would probably rather have the choice of physical model be the top level rather than the specifics of the mathematical formulation.

[benegee]

Thanks for your feedback @tristanmontoya !

Sounds good to me, and feels somewhat more natural to have the equations at the top level!

[benegee]

Here it is!
I also updated the description above.

I am still unsure about "cartesian_manifold". I think it makes sense to separate this from full 3D Cartesian?

[tristanmontoya]

While this is fine with me, what do you think of this modification?

• Keep the same top level
• Remove mid-level cartesian under euler until we have other options
• Make the options under shallow_water and advection into simply covariant and cartesian

[benegee]

Thanks again! @tristanmontoya

This make sense because we will not have different Cartesian options under shallow_water and advection, and we can spare one level.

It is also fine with me to not introduce further levels under euler until we have a covariant option.

[tristanmontoya]

Ok, this is good now. The only thing I see is that the example in the README.md should be updated to use the new directory structure as the currently referenced file doesn't exist. I don't care which case we use as the example one, main criteria is that it shouldn't take very long to run.