Interpolated CC state not consistent with interpolation class

[ZepeiX]

The binaries near the boundary might be classified as stable_MT or no_MT but not having interpolated CC properties for star_1.

[ZepeiX]

I also found that BHs from forward mass transfer can be classified as BH_reverse.

[astroJeff]

@ZepeiX , can you provide an example or two?

[astroJeff]

@maxbriel will also look into it. We need to check to see if it's still a problem

[maxbriel]

@astroJeff
Without altering the classifiers and the initial-final interpolator, these will remain at boundaries. There have been a few issues reported that show that it still exists in the classifiers.
For example:

• NS/WD NaN mass, SN_type & CO_type not consistent #280
• IF runs classifiers inconsistent #402
• incorrect assignment of "initial_RLOF" state #412

Besides the planned adjustment of the classifiers/interpolators, I don't think there's another fix possible at the moment. I would suggest we start to combine the issues into "one", if possible. That makes it easier to keep track of what issues it is causing.

[astroJeff]

Thanks for the update, @maxbriel. @philipp-rajah, this falls squarely in your wheelhouse. Have you had a chance to look into these issues?

[philipp-rajah]

@astroJeff I am looking into #402 and #280 right now. I haven't dug very deeply as of right now, but from a high level look it seems that these issues are a consequence of classifying these parameters independently. There will always be classification mismatches unless we constrain the classification somehow. Maybe this is something that should be discussed when we rewrite the interpolator.