maybe/2 would be a useful addition to library(random)

[triska]

A generalization of maybe/0 would be a useful addition to library(random):

maybe(+K, +N), succeeding with probability K/N.

This is needed for library(clpb), where we currently have:

?- sat(A+B), random_labeling(0, [A,B]).
   error(existence_error(procedure,maybe/2),maybe/2), unexpected.

@notoria and everyone who is interested in arithmetic and probabilistic functionality, I would greatly appreciate your help with this!

[UWN]

Note that there are systems (the original CLP(R) comes to mind) that do not produce as a full leaf answer containing constraints but instead just say maybe. Not sure this is such a good idea to repurpose this name.

[triska]

Yes indeed, very good point, a better name is also worth considering!

[hurufu]

Can this be enough?

maybe(K, N) :-
    maplist(integer_si, [K, N]),
    random_integer(0, N, X),
    X < K.

[triska]

Something in this direction would be perfect, if the bounds are correctly chosen to yield the correct probabilities!

[hurufu]

Also just a comment. For any serious security based application Prolog language should support multiple RNG instances for different purposes. For example if a program uses the same RNG to both generate access tokens and for UI animation, then it is much easier to predict next access tokens just by observing UI animation.

[bakaq]

Yes, ideally we need a way to explicitly generate and pass an "rng context" to rng functions so that we have control of this. It would also make things more declarative, depending on how the rng context works.