Bijectors.jl

Bijectors.jl copied to clipboard

CorrBijector makes posterior improper

7

An $n \times n$ correlation matrix has ${n \choose 2} = \frac{n (n-1)}{2}$ degrees of freedom. This is the same as the number of elements in a strict upper triangular...

sethaxen

MethodError: no method matching bijector(::MixtureModel{Multivariate, Continuous, MvNormal, Float64})

1

MWE: ```julia using Bijectors, Distributions dist = MixtureModel(MvNormal, [(ones(2), 1), (2 .* ones(2), 1)]) x = rand(dist) b = bijector(dist) ``` Error: ``` ERROR: MethodError: no method matching bijector(::MixtureModel{Multivariate, Continuous,...

krishvishal

This PR is an attempt at a couple of things: 1. Widening the scope, i.e. not being so restrictive regarding whether or not something is bijective. 2. Removal of expected...

torfjelde

Lightweight package for composed functions (function chains)

2

In the spirit of #199, which resulted in InverseFunctions and ChangesOfVariables: I've been looking for a lightweight package that defines composed functions that go beyond `Base.ComposedFunction`, so that allow for...

oschulz

Hi. I try to compile an app to predict bayesian model . I saved posterior chain, and model specification. My precomp_file is something like that. **precomp_file.jl** ```julia using Turing using...

joscani

This pull request changes the compat entry for the `Functors` package from `0.1, 0.2` to `0.1, 0.2, 0.3` for package test. This keeps the compat entries for earlier versions. Note:...

github-actions[bot]

This pull request changes the compat entry for the `Functors` package from `0.1, 0.2` to `0.1, 0.2, 0.3`. This keeps the compat entries for earlier versions. Note: I have not...

github-actions[bot]

`Bijectors.ordered(d)` is incorrect if `bijector(d)` is not `Identity`

1

`OrderedBijector` is only appropriate for bounded real numbers. It would be convenient to also have a `PositiveOrderedBijector` (similar to Stan's `positive_ordered` vector type), which wraps a distribution required to have...

sethaxen

I've found that Zygote fails to compute gradients when using the method of `logpdf` defined [here](https://github.com/TuringLang/Bijectors.jl/blob/master/src/transformed_distribution.jl#L93-L97) Here's a MWE: ```julia using Bijectors using DistributionsAD using Flux using Zygote d =...

tpgillam

Dear all, I wanted to ask why a `PDBijector` transforms the input to a lowertriangular matrix while a `CorrBijector` transforms the input into a uppertriangular matrix. It seems like most...

paschermayr