fernando

> with the ei function the point selected is never the one with highest utility function. > why? > > in any case if x[np.argmax(utility)] is -2 I would expect...

You can pass the `xi` parameter to make it explore more as seen in this [notebook](https://github.com/fmfn/BayesianOptimization/blob/master/examples/exploitation%20vs%20exploration.ipynb). I'll take a look at the code and try to get to the bottom...

The modification is explained [here](http://arxiv.org/pdf/1012.2599v1.pdf), its origin dates way back. And you are right, the real quantity of interest is xi/y_true_max, or something of the kind. You could alter the...

Yeah, I've noticed this too, I have some suspicions, but let me run some tests with the scripts you uploaded.

Hi @osullivryan , yes that sounds like a great idea. Given how time constraint I've been these days I won't be able to work on this feature anytime soon, but...

Thank you @Engineero , I will take a look at both. I'm a bit behind reviewing issues and PRs but will get around doing that soon.

My fear with implementing constraints on the acquisition function optimization is that the algorithm might get stuck probing the same edge points. Since the Gaussian Process is not constrained it...

Interesting, the kernel change they propose wouldn't be too hard to implement. My only concern is making the API more and more cumbersome by piling features. However this one is...

There's no way to exclude boundary points. I don't think the extra complexity would justify. As you mentioned you can simply use `(1e-4, 1)`, or something like it, since the...

Oh, I believe the way [colours are implemented](https://github.com/fmfn/BayesianOptimization/blob/af48b0374fc73851d34818d410b01e08eedb1e07/bayes_opt/util.py#L178) is not compatible with Windows. Unfortunately I don't have the bandwidth to fix this currently (nor do I have a way of...