David Widmann
David Widmann
Some things that probably can be improved as a follow-up to the initial implementation (but might require changes in other packages such as AD backends): - Avoid dense matrices (currently...
`Distributions.logpdf` is only defined for single samples (and not multiple samples) whereas `logpdf_with_trans` is defined for sets of samples as well. To be consistent with the behaviour in Distributions, one...
Currently, these distributions are broken due to https://github.com/FluxML/Zygote.jl/issues/873. As discussed in https://github.com/TuringLang/Bijectors.jl/pull/155, the corresponding tests are marked as broken. The issue will be fixed upstream.
It seems Distributions 0.24 broke ReverseDiff for the multivariate distribution `filldist(::Skellam, n)` (https://github.com/TuringLang/DistributionsAD.jl/pull/119#issuecomment-705769224).
This PR fixes https://github.com/JuliaStats/RDatasets.jl/issues/117 - and does not seem to break the tests. It specifies the file format explicitly when loading ".rda" files instead of relying on the automatic discovery...
Would it make sense to add the [Palmer penguin dataset](https://github.com/allisonhorst/palmerpenguins)? It was recently proposed as an alternative to the well-known Iris dataset due to growing sentiment about Ronald Fisher's eugenicist...
Currently, PriorityQueue is implemented as an AbstractDict by wrapping a `Dict` instance. Apparently, this design mixes two different data structures, and since this leads to performance issues (e.g., https://github.com/JuliaCollections/DataStructures.jl/issues/596) I'm...
This PR updates all tutorials with Turing 0.21 to StatsPlots 0.15 (tutorials with older Turing versions at least partly require more dependency changes and more careful inspection). The PR also...
I started to fix the PPCA tutorial. It's still work in progress but should already work for most parts, except the Householder example in the end (which I'll have to...
I'm mainly concerned about the, in most cases, incorrect results of `logprior` currently - which is the main reason for this annoying `leafcontext` approach. I think this could be fixed...