package prbnmcn-dagger-stats

  1. Overview
  2. Docs
Probabilistic programming library: prbnmcn-stats-based samplers

Install 

dune-project
 Dependency

github.com Readme Changelog MIT License Edit opam file Versions (6)

Authors

  1. I
    Ilias Garnier

Maintainers

  1. I
    igarnier@protonmail.com

Sources

0.0.3.tar.gz
md5=3fb82cc81ef81557f3325ef2c8611e35
sha512=41498634a1872cfd98afea4e35d8ffff08e444cd1a8dffd5d2a8aea57dc7578129c4832929fb6bf03c0f2edbfe1f85949533cf2e574273449ff526af1975aa28

doc/CHANGES.html

0.0.3

API changes

  • smc: provide API for programmable resampling (review by @nilsbecker, many thanks!)

  • gsl_dist: functorize over GSL signature

    • band-aid for unavailability of gsl package in ocaml 5
    • since explicit dependency on gsl is dropped, relicensed to MIT

code quality/bug fixes

  • smc: much improved documentation (report and review by @nilsbecker)
  • removed uses of Obj.magic

  • bugfix: properly handle terminated particles in smc

    • terminated particles now continue to participate in resampling events
  • bugfix: Smc.fork off-by-one corrected
  • bugfix: Cps_monad.map_array reversed order of elements

misc

  • added gamma distribution to stats_dist
  • added kalman filter example
  • added polynomial regression example
  • added approximate bayesian computation example
  • more tests

0.0.2

  • Dependency: prbnmcn-stats.0.0.3 -> prbnmcn-stats.0.0.4
  • Add beta distribution to Gsl samplers
  • Refactor Cps monad
  • Add SMC inference
  • Simplify handler type, modularize effect definitions away from Cps_monad
  • Fix typo: bernouilli -> bernoulli (report by @nilsbecker)

0.0.1

  • First release of prbnmcn-dagger, a library for probabilistic programming