gamlss is a R package implementing the Generalised Additive Models for The package gamlss
is an implementation of Rigby and Stasinopoulos (2005), Appl. Statist., 54, pp. 507-554.
There are three book available for information;
“Flexible Regression and Smoothing: Using GAMLSS in R” explaining how the models can be used in R.
“Distributions for modeling location, scale and shape: Using GAMLSS in R” explaining the explicit and generated distributions available in the package gamlss.dist
“Generized Additive Models for Location Scale and Shape: A distributional regression approach with applications” explaining the different method for fitting GAMLSS i.e. penalised Likelihood, Bayesian and Boosting.
More more information about books and papers related to GAMLSS can be found in https://www.gamlss.com/.
The GitHub repository is now hosted under the new gamlss-dev
organization: https://github.com/gamlss-dev/gamlss/.
predict()
do not print the message “new prediction”
stepGAIC()
produce less lines in the output
The package is now hosted on GitHub at https://github.com/gamlss-dev/gamlss/.
Add a new prodist()
method for extracting fitted (in-sample) or predicted (out-of-sample) probability distributions from gamlss models (contributed by Achim Zeileis). This enables the workflow from the distributions3 package for all distributions provided by gamlss.dist
. The idea is that the distributions3
objects encapsulate all information needed to obtain moments (mean, variance, etc.), probabilities, quantiles, etc. with a unified interface. See the useR! 2022 presentation by Zeileis, Lang, and Hayes for an overview.