External function optimization

[fipeop]

Hi, thanks for the great package! Very interesting new methods.

I wanted to check: what's the best way to deal with an external function? For example, I have data X and Y data from a black-box function (a physical experiment). I want to fit a model to that data and after an iteration of BO get the n best suggestion. These suggestions will later be evaluated in an external function, and I will update the values for X and Y.

For this I can fit a model with the data manually, then use the BO class with parallel evaluations it seems. Is there an easier way on an example for this case? How do I a list of suggestions after a first iteration of parallel BO?

[yarondav]

This looks really good! Thank you for the package.

I actually have the same question. I would love to understand how can I add other data points from a blackbox function. @fipeop - if you have further insights - I'd be interested.

[xmaww]

@AntKirill @fipeop The problem of multiple objective external function optimization also bothered me a long time，I found two basic tutorial here: ①https://nbviewer.org/github/SheffieldML/GPyOpt/blob/devel/manual/GPyOpt_external_objective_evaluation.ipynb ②https://github.com/fmfn/BayesianOptimization/blob/master/examples/advanced-tour.ipynb but both of them are single objective ones，do you have any breakthrough or any suggestions？Thanks a lot！