Failing test expected?

[Jamie Roberts]

Dear all,

On JEMRIS 2.9.1, I am receiving a failing test on JEMRIS when building from source (latest main branch, on ubuntu 22.04/wsl2), the failing test is one of the tests that is ran during Ctest -V after building with make, the test that fails is "Test Case 2: simulating MR signals" the radial.xml sanityck signal test

>>> 06.         radial.xml |        radial_signal (sig-simu)
>>> is NOT ok (e= 0.1803 %)

All other tests are passing. I'm exploring why this specific test is failing. I have also noted that on the pull request by brennerd11  submitted to JEMRIS that proposes to add a CI pipeline there is a specific modification to increase the tolerance for this test if failing to 0.2% thus allowing radial.xml to pass its sanitych. This leads me to suspect that this failing test is currently expected behaviour, or at least common.

Any ideas on what is causing it to fail? Should I be concerned? 

If anyone has any suggestions for the best way to investigate this do let me know.

I will admit I am a JEMRIS novice so hopefully this is not a silly question.

p.s. the update to JEMRIS is really helpful btw, in previous version of JEMRIS I found installing older CVODEs on newer ubuntu's a pain, this is much nicer, also great to not need MATLAB licence to run the GUI, much appreciation to Tony and those who have worked on it for the update.

Best regards

Jamie

[Artyom Tsanda]

Dear Jamie,

I'd like to echo your question as I have exactly the same issue (radial.xml test with 0.18%). As a novice, I cannot understand if it is critical or not. I'd be grateful for some support here from more experienced users/developers.

Best regards,
Artyom Tsanda

среда, 28 февраля 2024 г. в 09:39:40 UTC+1, Jamie Roberts:

[tony.s...@gmail.com]

Dear Jamie and Artyom,

Sorry for the late reply. The test result is not critical and potentially represents an issue with the test metrics. It's solved, but not yet pushed to the master branch on Github (I need to do a few other tests first).
There is a newer version in branch "mastertest" where all test cases should pass. However, that should not make any difference for using JEMRIS.

Cheers
Tony