Cbranincurrin synthetic test function for cmoo
#692
Conversation
| """ | ||
|
|
||
| def gen_pareto_optimal_points(n: int, seed: int | None = None) -> TensorType: | ||
| return tf.zeros(shape=0, dtype=tf.float64) |
There was a problem hiding this comment.
so there is no known pareto front for this problem I take it?
can you please add a note to the docs of the problem above stating that and that this function will return an empty tensor please
There was a problem hiding this comment.
@uri-granta do you think this is ok or we should return NotImplementedError? for some problems pareto front will not be available
it's a bit burdensome, but what can be done is to compute offline approximate ground truth with say genetic algos with a large number of function evaluations, store a fixed number of points in the repo as a text file and then load that when the function is called
There was a problem hiding this comment.
Definitely clearer to raise a NotImplementedError with a message explaining that there is no known pareto front, rather than returning an empty "front".
There was a problem hiding this comment.
@TsingQAQ can you please raise an error here, but perhaps in the docs you can indicate to user a way to generate approximate ground truth
There was a problem hiding this comment.
Sure! for readability have created a new exception class NoAnalyticalParetoPointsError, please check if this is idea.
There was a problem hiding this comment.
thanks @TsingQAQ, added a comment on your question on my suggestion about typing
and I think we need @uri-granta input on interfaces
3 participants
This is a (draft) PR for adding a commonly utilized synthetic test function for the multi-objective constraint function, which can be used in both notebooks and tests.
class ConstrainedMultiObjectiveTestProblem(MultiObjectiveTestProblem):as in this PR is the ideal approach. Alternatively, one can return a concatenation of objective function value and constraint in one function.