Sample paths to states in space within distance d from start state

[alisonparedes]

Sphere Stratified Sampling is a way to sample a uniform distribution of states from a domain's search space (unit cost) (Clauscker, SoCS-21).

The technique was designed to solve the problem that sampling in search spaces is usually biased away from the states that we care most about, but it incidentally also returns a set of paths used in the calculation.

See attached image.

[haz]

This seems eerily close to our goal sampling strategy! Means we must be on the right track...