AdvancedMH.jl
AdvancedMH.jl copied to clipboard
Robust implementation for random-walk Metropolis-Hastings algorithms
We are not currently tracking acceptances, which is a statistic that could be useful for manual tuning. We need to change the `Transition` struct to track accept/reject. In addition to...
I've implemented a simple version of the standard AM algorithm as a small extension to #39 (this PR is independent of #39). The pull request features a multivariate Gaussian proposal...
> Hi all. I recall that minus sign... Let me take a look. Yes perhaps within an issue is more appropriate. I will be a little slow though - I...
Acceptance status should be included in the transition struct for easy calculations of rejection rates.
Quasi-MH
I think that taking advantage of [this](https://www.pnas.org/content/102/25/8844) should let you get a significant improvement in sampling without being _too_ much work.
It is confusing that `Proposal`s such as `StaticProposal` accept arrays of proposals and one can also specify arrays of `Proposal`s in the `MetropolisHastings` sampler. Would it be sufficient to only...
I actually have one question, does AdvancedMH supports customized distributions?
Hi, I implemented some very basic adaptive Metropolis-Hastings proposals that may be of interest for the `AdvancedMH.jl` package. These are basically simple symmetric random-walk proposals where the scale is subjected...
I am quite new to Turing, but I really like how it composes with basically any Julia library which is fantastic. For hard problems, I think being able to customize...
Hi there, a question related to [this PR](#39) I'm currently facing an application where I would really like to use adaptive proposals like those defined in this PR in a...