Cornell-MOE
Cornell-MOE copied to clipboard
points_to_evaluate.shape
moe/optimal_learning/python/cpp_wrappers/expected_improvement_mcmc.py
class ExpectedImprovementMCMC
def evaluate_at_point_list()
line 219 says:
:type points_to_evaluate: array of float64 with shape (num_to_evaluate, self.dim)
so the shape here should be a tuple of 2.
but in line 246: num_to_evaluate, num_to_sample, _ = points_to_evaluate.shape it's expected tuple of 3.
Is there something wrong?
(1) All the comments are outdated. We will put significant efforts to revise all the comments this summer, you can expect an update then. (2) evaluate_at_point_list() is not tested, and subject to errors, instead please use compute_objective_function() within a loop over the point list.
thanks, that works.