Application fixes

[akeeste]

Tied to and should be merged after WEC-Sim/WEC-Sim#1401.
Fixes WEC-Sim/WEC-Sim#1400

[akeeste]

@jtgrasb the moordyn test is now passing here too! The mooring matrix case is what fails the mooring test now

[akeeste]

I ultimately chose to fix the passive yaw test by replacing the results that our regression test compares against because:

• the passive yaw results are identical when using the previous linear interpolation instead of the current spline interpolation.
• the current difference in results is extremely small. The regression test currently only allows a zero tolerance, which is likely more stringent than is warranted.
• The spline vs linear interpolation in bodyClass.irrExcitation is the only source of difference, and we could argue this method is more accurate to interpolate against anyways

@dforbush2 Let me know if you disagree with this approach

[dforbush2]

I do not disagree. I opted for spline in this case because it allows the specification of the extrapolated value (or something like that?) that linear does not...and I would like to have the ability to consider a fixed/zero'd out set of hydrodynamic data beyond a certain extent. This facilitates runs where the range of motion is known a priori by making the required range of BEM data smaller. From: Adam Keester ***@***.***> Sent: Thursday, January 23, 2025 2:32 PM To: WEC-Sim/WEC-Sim_Applications ***@***.***> Cc: Forbush, Dominic Dean ***@***.***>; Mention ***@***.***> Subject: [EXTERNAL] Re: [WEC-Sim/WEC-Sim_Applications] Application fixes (PR #80) I ultimately chose to fix the passive yaw test by replacing the results that our regression test compares against because: * the passive yaw results are identical when using the previous linear interpolation instead of the current spline interpolation. * the current difference in results is extremely small. The regression test currently only allows a zero tolerance, which is likely more stringent than is warranted. * The spline vs linear interpolation in bodyClass.irrExcitation is the only source of difference, and we could argue this method is more accurate to interpolate against anyways @dforbush2<https://github.com/dforbush2> Let me know if you disagree with this approach - Reply to this email directly, view it on GitHub<#80 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AHF2G7HZW5D3XPI6FBWOGK32MFNT7AVCNFSM6AAAAABVKKLMD6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMJRGA2TQMJUGU>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[akeeste]

@kmruehl assigning you for now since WEC-Sim/WEC-Sim#1401 is also assigned to you and it's for the same effort to resolve application bugs

[akeeste]

@kmruehl this is ready for review

[akeeste]

@kmruehl any updates on the review for this PR?

[kmruehl]

This PR looks good. I revised the tests to only test the R2024b version of MATLAB, and I'm waiting for the tests to pass to merge.

[kmruehl]

@akeeste Mooring and Paraview tests are not yet passing, please work with @dforbush2 to resolve while I'm out.

[akeeste]

The moordyn tests (mooring/moordyn and paraview/RM3) both fail sporadically, crashing MATLAB. The error is unintelligible and cannot be produced. Tests pass locally so I propose we just remove them and revisit those tests at a different time.

The Paraview/OSWEC accelerator build fails for some reason. I will dig into this again and hopefully get these tests finally working

[akeeste]

@kmruehl what was the problem with the Paraview OSWEC slx file? It passed on R2023b but is failing now on R2024b. I get an warning that blocks were removed when it was exported from R2024a to R2023b (though it doesn't end up showing any for which that was the case). I don't have R2024b to test with.

[akeeste]

@dforbush2 @kmruehl The runner is not working robustly. Paraview OSWEC works locally with R2024b. I propose we skip both paraview and moordyn tests for now so that we can finish the progress made here and move on.

[kmruehl]

@akeeste I didn't change anything within the ParView OSWEC *.slx file, I only saved it with a newer version of MATLAB, since we were no longer testing the version of MATLAB.

[kmruehl]

@akeeste I'm okay with removing those tests now if you've confirmed that they pass locally, and the issue is with the runner.

[akeeste]

@kmruehl @dforbush2 I confirmed again this afternoon that all 3 of the breaking cases (Mooring/MoorDyn, Paraview OSWEC, Paraview RM3 MoorDyn) cases work locally with R2024b. I cannot easily separate the MooringMatrix test from the MoorDyn test so I skipped it too.

But in testing the MooringMatrix case locally I found that it is now not working🙃. It runs fine on normal and accelerator modes, but not rapid-accelerator. These bugs never end...

We could skip that one too--that would be easiest regardless, but we should be aware there may be problems with it.

[kmruehl]

@akeeste I'm going to go ahead and merge this now to get the interim bug-fixes into main, however we can you pleas update the related issue to note the issues with MooringMatrix?

Thank you

[akeeste]

FYI that MooringMatrix with rapid-accelerator did work locally (windows) with R2024a. rapid-accelerator issue might be specific to R2024b.