SNe Stellar_Type in SSE equal to Stellar_Type_Prev instead of remnant type

[pauldisberg]

Description:
In SSE, the Stellar_Type results in Supernovae describe the progenitors and not the formed compact objects. Stellar_Type is equal to Stellar_Type_Prev. The Stellar_Type results in System_Patameters, as well as Stellar_Type in BSE mode, do give the (desired) compact object stellar types.

To Reproduce:
Run ./COMPAS -n 10 --mode SSE --random-seed 1714388113
Access h5.File(COMPAS_Outputs.h5)['SSE_Supernovae']['Stellar_Type'][()]

Versioning:

• OS: Ubuntu 24.04.2 LTS
• COMPAS v03.20.02

[ilyamandel]

Looks like SSE and BSE are not behaving consistently in recording the Stellar_Type output... @jeffriley ? :)

[SimonStevenson]

[I just rediscovered this bug whilst testing #1404]

[jeffriley]

This happens because we switch stellar type at the end of the timestep in SSE, but we print the SN file record when the SN happens (mid-timestep). The problem doesn't occur in BSE mode because we forrce the stellar type switch (either in the envelope loss code or the mass loss/transfer code). In the short term we could fix this by moving the SN file print to the end of the timestep in SSE mode - but then all the attributes of the star would be the values at the end of the timestep, not the mid-timestep values at the time of the SN. Ultimately this relates to issue #805 - we should switch and end the timestep immediately the timestep occurs.

So - should we move the SN print to the end of the timestep for now, or just leave this until I fix issue #805 ? Happy to take suggestions.

EDIT: this may have just been noticed because we removed the "delayed" SN print when we removed the revert code for SSE. I can't remember if the delayed SN print picked up the attributes mid-timestep, or at the end of the timestep when the print was actually done (I could look I suppose...).

[ilyamandel]

Do you have an estimate for when #805 might happen, @jeffriley ? If the future is not too distant, it probably makes sense to wait rather than apply piecemeal patches...

[jeffriley]

@ilyamandel, every time I think about #805 I think probably a couple of weeks - then I look at the code :-) Realistically, I think I can probably have something by the end of July - I think I have worked out what needs to be done, and now just have to work out how best to implement it.

[ilyamandel]

Then I'd say let's wait for another month on this one.