[R] General partition search algorithm (local search, genetic algorithm, ...)

[Aleš Žiberna]

Dear R users!

I was wondering if there exists (in R) any general algorithm for finding
optimal partition (optimal allocation of n units into k groups or bins),
such as local search, genetic algorithm, tabu search, ...

By general I mean such that would find an (approximately) optimal partition
based on some user specified criterion function, that would be evaluated on
the data and partition.

It is especially essential that the algorithm does not require any unit
fitness values, only the value of criterion function (or fitness) for the
partition as a whole.

Thanks for any replies!

Best regards,
Ales Ziberna