https://github.com/tomicapretto/shiny-hmc

A Shiny app to help people learn (and play with) Hamiltonian Monte Carlo
https://github.com/tomicapretto/shiny-hmc

bayesian-statistics hmc mcmc

Last synced: 2 months ago
JSON representation

A Shiny app to help people learn (and play with) Hamiltonian Monte Carlo

• Host: GitHub
• URL: https://github.com/tomicapretto/shiny-hmc
• Owner: tomicapretto
• License: apache-2.0
• Created: 2024-04-10T23:36:21.000Z (9 months ago)
• Default Branch: main
• Last Pushed: 2024-04-27T15:04:46.000Z (8 months ago)
• Last Synced: 2024-04-27T16:24:07.594Z (8 months ago)
• Topics: bayesian-statistics, hmc, mcmc
• Language: R
• Homepage:
• Size: 1.62 MB
• Stars: 3
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE.md

Awesome Lists containing this project

README

        
# ShinyHMC

An interactive Shiny app to learn about Hamiltonian Monte Carlo sampling.

This is still a work in progress. I do have a couple of ideas to try and I would like to document everything a bit more. Suggestions are more than welcomed :)

## Installation

Install this package from [GitHub](https://github.com/) with:

``` r

# install.packages("devtools")

devtools::install_github("tomicapretto/shiny-hmc")

```

## Run application

The function `run_app()` fires up the application. Just do

``` r

shinyhmc::run_app()

```

and have fun!

## Ideas

### Features

* Add content about each target distribution in the main container

* Change color of segments as more segments are added?

  * To make clear when a trajectory returns over the same point

  * Gradient instead of transparency?

* Allow users to specify a momentum

### Usage

* Show what happens when momentum is null (i.e. c(0, 0))

* Show what happens when momentum is huge (e.g. c(4, 3.5))

* Show what happens when momentum is exactly the same in both directions (c(2, 2))