Evaluation of acquisition function in the regular BO loop

[knstntn-ts]

There seems to be a typo in the implementation of the regular BO loops. Similar to the Issue #2, X values where an acquisition function is evaluated might need to be changed. In the following line:

ZoMBI/zombi.py

Line 149 in 04a7f0f

ac_value = self.BO(X=X_new, GP_model=self.GP, n=n, fX_best=fX_best, fX_best_min=fX_best_min,

self.BO is calculated for X_new, but I think it should be evaluated on all available X values, i.e., self.dataset_X.

Attached is the notebook where I compared the performance of the two implementations. It seems that when X_new is replaced with self.dataset_X the performance slightly improves.

zombi_test_2.zip