README.md

In kleinschmidt/beliefupdatr: Belief updating for phonetic adaptation

Belief UpdatR

Belief updating for phonetic adaptation in R.

Install

Prerequisites

Parts of this package depend on having a working developer environment. You may already have these, but if not:

• Mac OS: install XCode (from the App Store)
• Windows: install RTools
• Linux: you probably know what you're doing.

The RStan installation instructions walk through this process in more detail.

Finally, install devtools if it's not already installed:

install.packages('devtools')

From github

Second, install this package from github.

devtools::install_github('kleinschmidt/beliefupdatr', args = "--preclean")

If you want to build the vignettes (long-form documentation), there are additional dependencies. To install this package with optional dependencies (Suggests in the DESCRIPTION file) and build the vignettes:

devtools::install_github('kleinschmidt/beliefupdatr', args = "--preclean",
                         dependencies=TRUE, build_vignettes=TRUE)

This will take a bit longer, because the vignettes run some simulations.

Getting help

Functions are documented in the usual way:

help(package='beliefupdatr')
?beliefupdatr::d_nix2_predict

# or
library(beliefupdatr)
?d_nix2_predict

Long-form documentation is available in the form of vignettes:

vignette(package='beliefupdatr')
vignette('conjugate-updating')

Or to open a list of vignettes in your web browser:

browseVignettes(package='beliefupdatr')

Contents

• /exec has source for Stan models. You'll need rstan to run them. Construct the full path to the model file like file.path(devtools::inst('beliefupdatr'), 'exec', '<model.stan>').
• /R/nix.R: helper functions for the Normal-Chi^-2 distribution which is a conjugate prior on a Normal distribution with unknown mean and variance.

kleinschmidt/beliefupdatr documentation built on May 24, 2020, 8:26 p.m.