David Widmann

`NewtonTrustRegion` algorithm deviates from reference

1

When looking through the code of `NewtonTrustRegion`, I noticed that the implementation seems to differ a bit from the algorithm 4.1 in N&W that it refers to. Is there an...

Fix implementations of `eigen` and `eigvals`

2

The PR adds missing definitions of `eigvals` and `eigen` for `Hermitian{ @time test_hessian(Float64) 0.000071 seconds (38 allocations: 33.750 KiB) 9×9 Matrix{Float64}: 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0...

Bump version from 0.25.122 to 0.25.123

1

To release https://github.com/JuliaStats/Distributions.jl/pull/1930. Fixes #2018.

Add log-logistic distribution

1

Updated and improved version of #1454. Adding reference tests against R is a bit cumbersome currently, see #2007. Edit: I added tests against R nevertheless by locally commenting out all...

Use `Rmath` instead of private `StatsFuns.RFunctions` submodule

1

Currently, Distributions uses non-public functions in the private submodule `StatsFuns.RFunctions`. This could break in any new release of StatsFuns. This PR switches the code to public functions in `Rmath` (which...

Fix R test references and make them reproducible

1

When working on adding a new distribution, I realized (once again) that the R test references can't be re-generated currently since running `Rscript gendref.R` errors: - The `LindleyR` package is...

`force_latest_compatible_version=true` tries to use the latest version of the dependencies, irrespective of whether they are compatible with the julia version. This is problematic for packages such as JET where latest...

Builds on #256 and adds tests with Aqua. To make them pass, it was necessary to fix a handful of method ambiguity errors. Edit: It seems there are additional ambiguity...

The whole parent array is serialized/deserialized when sending around `SubArray`s between workers.