Zenna Tavares

- [x] Add RCD operator - [ ] Add RCD tests - [x] Add RID operator - [ ] Add RID tests - [ ] sat example In old Omega...

Extend lifting distributions to beyond Normal / Uniform

1

Currently, we can do: ``` \mu = @~ Normal(0, 1) X = Normal(\mu, 1) ``` and it will create the corresponding pointwise application, but it is does not work beyond...

``` μ = @~ Normal(0, 1) x = @~ Normal(μ, 1) randsample((μ, x)) x_ = 0.1234 μ_posterior = μ |ᶜ x ==ₚ x_ ``` Sampling from `μ_posterior` produces false. This...

Currently, in #lang, we are using integer and integer like types for indices. It makes sense to broaden these to basically any type, e.g.: ``` :name ~ Normal(0, 1) ```...

Echoing https://www.youtube.com/watch?v=_wfZSvasLFk, implement distributions which are explicitly parameterized as points on the quantile/CD-R. One idea is to use function classes which are guaranteed to be monotonicm e,g, some classes of...

Implement `grad(x, ω, ForwardDiff)`

1

Implement gradient using ForwardDiff (and/or reverseDiff)

- [ ] (non silent) error on attempting to intervene unintervenable - [ ] choose syntax 1. Y | had(X => x) 2. Y ¦ (X => x) -- concise,...

If `f` is some random variable, we should support `f(rng)` where `rng` is an ordinary Random.AbstractRNG object, such as a mersennte twister

Memoization is important in Omega because shared parentage means its easily to do redundant computation - [ ] Type safe - [ ] memoize interacts - [ ] Tests