Inspired by cats.Parallel, for modeling e.g. the relationship between Double and LogDouble. They are the same datatype, with two different algebras.
Practically, this helps to interface between probability-native APIs (where LogDouble is an ideal representation) and "numerical" APIs (typically inference algorithms, which may be in terms of e.g. potential energy).
Inspired by
cats.Parallel, for modeling e.g. the relationship betweenDoubleandLogDouble. They are the same datatype, with two different algebras.Practically, this helps to interface between probability-native APIs (where
LogDoubleis an ideal representation) and "numerical" APIs (typically inference algorithms, which may be in terms of e.g. potential energy).