Update MvNormal

[cscherrer]

This PR makes simplifications to MvNormal, based on the new Affine constructor in MeasureBase

[codecov]

Codecov Report

Merging #160 (7ce5d69) into master (0890e24) will increase coverage by 5.78%.
The diff coverage is 67.74%.

❗ Current head 7ce5d69 differs from pull request most recent head 90a3154. Consider uploading reports for the commit 90a3154 to get more accurate results

@@            Coverage Diff             @@
##           master     #160      +/-   ##
==========================================
+ Coverage   37.26%   43.05%   +5.78%     
==========================================
  Files          29       30       +1     
  Lines         585      583       -2     
==========================================
+ Hits          218      251      +33     
+ Misses        367      332      -35     

Impacted Files	Coverage Δ
src/combinators/affine.jl	0.00% <0.00%> (ø)
src/combinators/product.jl	0.00% <0.00%> (ø)
src/parameterized.jl	72.72% <ø> (+22.72%)	⬆️
src/parameterized/gumbel.jl	18.18% <25.00%> (+5.68%)	⬆️
src/MeasureTheory.jl	50.00% <50.00%> (+40.00%)	⬆️
src/parameterized/studentt.jl	70.83% <72.72%> (+9.72%)	⬆️
src/parameterized/normal.jl	38.23% <90.90%> (+14.09%)	⬆️
src/distproxy.jl	50.00% <100.00%> (ø)
src/parameterized/cauchy.jl	62.50% <100.00%> (+31.73%)	⬆️
src/parameterized/laplace.jl	69.23% <100.00%> (+59.23%)	⬆️
... and 4 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 0890e24...90a3154. Read the comment docs.

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[cscherrer]

I think it's good! This test passes:

@testset "MvNormal" begin
    Q,R = qr(randn(4,2))
    # Make sure `R` has a positive diagonal
    D = Diagonal(sign.(diag(R)))
    Q = Matrix(Q) * D
    R = D * R

    z = randn(2)
    
    ℓ = logpdf(MvNormal((σ= R,)),z)
    @test ℓ ≈ Dists.logpdf(Dists.MvNormal(R*R'),z)
    @test ℓ ≈ logpdf(MvNormal((σ= Q*R,)),Q*z)
end

But I broke a bunch of rand! stuff along the way, so I need to fix that before merging

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[cscherrer]

Hi @mschauer, Breakage shows Mitosis tests failing, but it looks like it's just a version bound issue. Can you have a look? This is part of a fix for a bug that was causing a big problem in Soss (cscherrer/Soss.jl#305) so I'm hoping to get it merged soon. If you don't get a chance today I'll go ahead and merge it as a major release.

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl

[github-actions]

Package name	latest	stable
Mitosis.jl
Soss.jl