Wish list

[cscherrer]

Measures from scratch

• [x] Specify equivalence class (wrt absolute continuity)
• [x] Density
• [ ] Probability measures
• [ ] Random sampling
• [x] Parameterized measures
• [ ] Measure of a set (for some extensible definition of "set")

Composition of Measures

• [x] Superposition
• [x] Scaling
• [ ] Product measure
• [ ] Composition of parameterized measures ("mixture measures", formally a monadic bind)
• [ ] Reparameterization
• [ ] Push-forwards (or is it "pushes-forward"?)

Extensibility

• [ ] Easily make measures from other packages a Measure

Computations

• [ ] Check for absolute continuity
• [ ] Radon-Nikodym derivatives
• [ ] Lebesgue decomposition

[sethaxen]

Many of these I don't know much about, but I'm very excited about the ones I do. Where would changing the measure fall in this list? e.g. One has a density wrt one measure and wants to express the density wrt a new measure. Reparameterization?

[cscherrer]

Sorry I had missed this @sethaxen . Yes, I think that's right. We'll have an \int operator, and it can fall back to a Integral struct. I've added some thoughts in the README on this :)