Latin hypercube strategy needs to know the total iteration count up-front but it cannot get this information. The temporary solution in #96 is to introduce a separate n_max parameter for the Latin strategy, but there is little use in this having a different value.
So the proposal is to add n to the signature of setup:
state = setup(tuning::MyTuningStrategy, model, range, n, verbosity)
Latin hypercube strategy needs to know the total iteration count up-front but it cannot get this information. The temporary solution in #96 is to introduce a separate
n_max
parameter for the Latin strategy, but there is little use in this having a different value.So the proposal is to add
n
to the signature ofsetup
: