https://github.com/lilithhafner/aliastables.jl

An efficient sampler for discrete random variables
https://github.com/lilithhafner/aliastables.jl

Last synced: 2 months ago
JSON representation

An efficient sampler for discrete random variables

• Host: GitHub
• URL: https://github.com/lilithhafner/aliastables.jl
• Owner: LilithHafner
• License: mit
• Created: 2024-03-25T01:20:14.000Z (about 1 year ago)
• Default Branch: main
• Last Pushed: 2024-11-18T00:52:34.000Z (6 months ago)
• Last Synced: 2025-02-20T21:41:20.296Z (3 months ago)
• Language: Julia
• Size: 2.33 MB
• Stars: 10
• Watchers: 2
• Forks: 1
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# AliasTables

[![docs](https://img.shields.io/badge/docs-v1-blue.svg)](https://aliastables.lilithhafner.com/dev)

[![Build Status](https://github.com/LilithHafner/AliasTables.jl/actions/workflows/CI.yml/badge.svg?branch=main)](https://github.com/LilithHafner/AliasTables.jl/actions/workflows/CI.yml?query=branch%3Amain)

[![Coverage](https://codecov.io/gh/LilithHafner/AliasTables.jl/branch/main/graph/badge.svg)](https://codecov.io/gh/LilithHafner/AliasTables.jl)

[![PkgEval](https://JuliaCI.github.io/NanosoldierReports/pkgeval_badges/A/AliasTables.svg)](https://JuliaCI.github.io/NanosoldierReports/pkgeval_badges/A/AliasTables.html)

[![Aqua](https://raw.githubusercontent.com/JuliaTesting/Aqua.jl/master/badge.svg)](https://github.com/JuliaTesting/Aqua.jl)

[![deps](https://juliahub.com/docs/AliasTables/deps.svg)](https://juliahub.com/ui/Packages/AliasTables/6dCmw?t=2)

AliasTables provides the `AliasTable` type, which is an object that defines a probability

distribution over `1:n` for some `n`. They are efficient to construct and very efficient to

sample from.

An alias table can be combined with a dense vector of values to create a discrete

distribution over anything.

Internally, AliasTables define a mapping from an unsigned integer type to the sampling

domain. To get a random sample according to the AliasTable's distribution, one must provide

a random unsigned integer uniformly at random. One can also provide a `Random.AbstractRNG`

object instead and a random unsigned integer will be generated using that rng. When using

the random API, this latter approach is taken.

```julia

julia> using AliasTables

julia> at = AliasTable([5,10,1])

AliasTable([0x5000000000000000, 0xa000000000000000, 0x1000000000000000])

julia> rand(at, 10)

10-element Vector{Int64}:

 2

 1

 2

 2

 2

 2

 1

 1

 3

 2

julia> using Chairmarks

julia> @b at rand

2.898 ns

julia> @b rand(UInt)

2.738 ns

julia> @b rand(1000) AliasTable

9.167 μs (2 allocs: 16.031 KiB)

julia> @b AliasTable(rand(1000)) rand(_, 1000)

1.506 μs (3 allocs: 7.875 KiB)

julia> @b AliasTable(rand(1000)), rand(1000) AliasTables.set_weights!(_...)

8.427 μs

julia> using StatsBase

julia> at = AliasTable{UInt16}([5,10,1])

AliasTable{UInt16}([0x5000, 0xa000, 0x1000])

julia> countmap(AliasTables.sample(x, at) for x in typemin(UInt16):typemax(UInt16))

Dict{Any, Int64} with 3 entries:

  2 => 40960

  3 => 4096

  1 => 20480

```