Towards partitioned RHS's (1D)

DanielDoehring commented 3 weeks ago

This draft is one way how we could realize partitioned RHS's with relatively little changes to the existing code. Opinions on this @sloede ?

github-actions[bot] commented 3 weeks ago

Review checklist

This checklist is meant to assist creators of PRs (to let them know what reviewers will typically look for) and reviewers (to guide them in a structured review process). Items do not need to be checked explicitly for a PR to be eligible for merging.

Purpose and scope

[ ] The PR has a single goal that is clear from the PR title and/or description.
[ ] All code changes represent a single set of modifications that logically belong together.
[ ] No more than 500 lines of code are changed or there is no obvious way to split the PR into multiple PRs.

Code quality

[ ] The code can be understood easily.
[ ] Newly introduced names for variables etc. are self-descriptive and consistent with existing naming conventions.
[ ] There are no redundancies that can be removed by simple modularization/refactoring.
[ ] There are no leftover debug statements or commented code sections.
[ ] The code adheres to our conventions and style guide, and to the Julia guidelines.

Documentation

[ ] New functions and types are documented with a docstring or top-level comment.
[ ] Relevant publications are referenced in docstrings (see example for formatting).
[ ] Inline comments are used to document longer or unusual code sections.
[ ] Comments describe intent ("why?") and not just functionality ("what?").
[ ] If the PR introduces a significant change or new feature, it is documented in NEWS.md with its PR number.

Testing

[ ] The PR passes all tests.
[ ] New or modified lines of code are covered by tests.
[ ] New or modified tests run in less then 10 seconds.

Performance

[ ] There are no type instabilities or memory allocations in performance-critical parts.
[ ] If the PR intent is to improve performance, before/after time measurements are posted in the PR.

Verification

[ ] The correctness of the code was verified using appropriate tests.
[ ] If new equations/methods are added, a convergence test has been run and the results are posted in the PR.

Created with :heart: by the Trixi.jl community.

codecov[bot] commented 3 weeks ago

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 96.16%. Comparing base (def208b) to head (78a7786).

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #1967 +/- ## ======================================= Coverage 96.16% 96.16% ======================================= Files 460 460 Lines 36980 36980 ======================================= Hits 35560 35560 Misses 1420 1420 ``` | [Flag](https://app.codecov.io/gh/trixi-framework/Trixi.jl/pull/1967/flags?src=pr&el=flags&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=trixi-framework) | Coverage Δ | | |---|---|---| | [unittests](https://app.codecov.io/gh/trixi-framework/Trixi.jl/pull/1967/flags?src=pr&el=flag&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=trixi-framework) | `96.16% <100.00%> (ø)` | | Flags with carried forward coverage won't be shown. [Click here](https://docs.codecov.io/docs/carryforward-flags?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=trixi-framework#carryforward-flags-in-the-pull-request-comment) to find out more.

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

sloede commented 1 week ago

This draft is one way how we could realize partitioned RHS's with relatively little changes to the existing code. Opinions on this @sloede ?

From a first look, I like the idea!

An small variation could be to have a modified function without default argument and create new functions with the old API that call it with eachelement(...) etc. But I don't know if that's really more readable. Have you checked if your change impacts the performance (since now maybe the compiler is not able to infer as much as before)?

Another thought is that to me, range strongly implies an ordered subset of an array, e.g., something like 5:10 or 10:2:20. However, in this case there is no such a restriction - it is just a bunch of indices. So maybe {element,interface,whatever}_indices would be more apt?

sloede commented 1 week ago

In general, I think this would be a good topic to bring up at a Trixi.jl meeting for discussion (possibly with a heads up in Slack such that people can think about it beforehand)

DanielDoehring commented 1 week ago

An small variation could be to have a modified function without default argument and create new functions with the old API that call it with eachelement(...) etc. But I don't know if that's really more readable.

True, but this adds in principle the overhead of calling another function, right? But should be explored in a benchmarking run.

Have you checked if your change impacts the performance (since now maybe the compiler is not able to infer as much as before)?

No, not yet.

Another thought is that to me, range strongly implies an ordered subset of an array, e.g., something like 5:10 or 10:2:20. However, in this case there is no such a restriction - it is just a bunch of indices. So maybe {element,interface,whatever}_indices would be more apt?

Yeah that sounds reasonable :+1:

sloede commented 1 week ago

An small variation could be to have a modified function without default argument and create new functions with the old API that call it with eachelement(...) etc. But I don't know if that's really more readable.

True, but this adds in principle the overhead of calling another function, right? But should be explored in a benchmarking run.

Yes and yes. In the end, it is likely that it doesn't make a difference performance-wise.

ranocha commented 1 week ago

Sounds reasonable. Could you please run some benchmarks?

DanielDoehring commented 5 days ago

Currently, we do not benchmark any 1D simulation:

https://github.com/trixi-framework/Trixi.jl/blob/961f2e7423bd14c6d64170c296ea8238a26a2995/benchmark/benchmarks.jl#L13-L39

I could add some 1D elixirs locally or we extend the benchmarks by some representative elixirs for 1D.

Otherwise, I could extend these changes to 2D to get benchmark results for this.

ranocha commented 4 days ago

I think it would be nice if you could add some representative 1D elixirs to the benchmarks in a new PR

trixi-framework / Trixi.jl