Pedro Bruel

If the values of options when Type is A are completely independent of the values of options when Type is B, it seems that you have two separate search spaces....

Ok, so it seems it would really make more sense to tune each type separately, specially if you can't rule out interactions between type-independent and type-specific options. An argument for...

You would have to build your dictionary using a `for` loop, with a `FloatParameter` for each of your 6000 parameters

I hadn't seen it, thanks for posting. Would you be interested in forking this project?

Thanks for calling attention to this. I don't have access to this repository, hence the creation of the fork. I do not oppose to deleting `phrb/ob-julia` and replaying any changes...

Thanks for the openness @gjkerns! I'm maintaing `ob-julia` now. Would you be willing to transfer ownership of this repository to ~me~ someone, such as @tarsius, so I could keep track...

By all means, it does not have to be transferred to me, I'm just taking initiative :smiley: I'm happy to contribute changes to someone else's repository.

I'm happy to help developing `ob-julia` wherever it ends up. I had volunteered to maintain it in a thread at the org-mode list where @bzg asked for volunteers. Since apparently...

Hey, that looks nice, especially your implementations of federov's and K-exchange! I've implemented KL-exchange (https://github.com/phrb/ExperimentalDesign.jl/blob/master/src/kl_exchange.jl#L109), although not so cleanly as you did. I tried to use matrix and vector views...

A modification may not be necessary if #24 is implemented correctly.