Seth Axen
Seth Axen
Once https://github.com/baggepinnen/MonteCarloMeasurements.jl/pull/58 is merged, the way will be paved to make Soss's calls to `particles` and `rand` optionally take a first argument `rng::AbstractRNG`. This will make Soss runs more reproducible...
`rand_phasepoint` should pass the point `q` to `rand`. i.e. https://github.com/tpapp/DynamicHMC.jl/blob/65f8a30ed0cc74c26206dcbbdc6dee91629dfddf/src/hamiltonian.jl#L157 should be ```julia rand_phasepoint(rng::AbstractRNG, H, q) = phasepoint_in(H, q, rand(rng, H.κ, q)) ``` This is consistent with the existing interface...
It would be nice to early on adopt an existing style guide so the conventions are well established. I use Invenia's [Blue Style](https://github.com/invenia/BlueStyle) for my repos. JuliaFormatter.jl's defaults match Blue...
Has there been discussion on if/how fitting of measures from draws should be implemented? I'm also wondering if/how estimators of moments from draws or exact moments when known should be...
If we implement the `Random.Sampler` interface, then we can draw IID samples from the distributions: https://docs.julialang.org/en/v1/stdlib/Random/#Generating-random-values-of-custom-types. Distributions tends to make scalar and array distributions return `Array`s, but should we instead...
Fixes #128
In https://github.com/JuliaManifolds/ManifoldMeasures.jl/issues/7 we were discussing that we need a way to map an (unnormalized) Hausdorff measure to the corresponding probability measure by normalizing. This seems general enough that it could...
I see that InfiniteArrays only works for Julia 1.5. Is this dependency required, or can it be conditionally loaded with Requires so offer support for older Julia versions? Are there...
I came across some issues with `MeasureTheory.LKJL` while trying to use it with Soss. I'm posting the issue here because I'm not 100% sure which issues need to be resolved...
This package defines `×(μ::AbstractMeasure{X}, ν::AbstractMeasure{Y})`. This will collide with `LinearAlgebra.cross`, which is aliased to `×`, which is exported. The solution we used in Manifolds.jl was to overload `LinearAlgebra.cross` and then...