ID for candidate

[martinhaas]

Hello,

This is related to the repeatable results question posed a couple of weeks ago. I need someway of identifying a Candidate with a repeatable ID. For example, candidate #27 has the same chromosome and produces the result every time I run the optimization. Should I add a static int ID to the candidate class and increment every time a candidate is produced? I do have access to the thread ID, but this is from within the pool (never goes larger than than the number of execution_threads. If I'm running 200 hundred generations, each with a population of 100, I would need IDs between 1 and 20,000. Any thoughts about where to put this information? Maybe something already exists and I'm overlooking it. Thanks again

[KRM7]

What is the ID needed for? Depending on the use case, the IDs might not need to be a part of the candidates.

The library currently doesn't include unique IDs for the candidates or anything that could be used as such. The closest thing is that you can hash the candidates using std::hash, but this obviously won't guarantee that different candidates won't hash to the same value. When it comes to identifying the candidates, their chromosomes are ultimately what identifies them.