glmmTMB balances speed and flexibility among packages for Zero-inflated Generalized Linear Mixed Modeling

Mollie Elizabeth Brooks, Kasper Kristensen, Koen J. van Benthem, Arni Magnusson, Casper Willestofte Berg, Anders Nielsen, Hans J. Skaug, Martin Machler, Benjamin M. Bolker

Research output: Contribution to journalJournal articleResearchpeer-review

795 Downloads (Pure)

Abstract

Count data can be analyzed using generalized linear mixed models when observations are correlated in ways that require random effects. However, count data are often zero-inflated, containing more zeros than would be expected from the typical error distributions. We present a new package, glmmTMB, and compare it to other R packages that fit zero-inflated mixed models. The glmmTMB package fits many types of GLMMs and extensions, including models with continuously distributed responses, but here we focus on count responses. glmmTMB is faster than glmmADMB, MCMCglmm, and brms, and more flexible than INLA and mgcv for zero-inflated modeling. One unique feature of glmmTMB (among packages that fit zero-inflated mixed models) is its ability to estimate the Conway-Maxwell-Poisson distribution parameterized by the mean. Overall, its most appealing features for new users may be the combination of speed, flexibility, and its interface's similarity to lme4.
Original languageEnglish
JournalThe R Journal
Volume9
Issue number2
Pages (from-to)378-400
ISSN2073-4859
DOIs
Publication statusPublished - 2017

Keywords

  • COMPUTER
  • STATISTICS
  • MAXWELL-POISSON DISTRIBUTION
  • COUNT DATA
  • BAYESIAN-INFERENCE
  • R PACKAGE
  • REGRESSION
  • ECOLOGY
  • EVOLUTION
  • ABUNDANCE

Cite this

Brooks, Mollie Elizabeth ; Kristensen, Kasper ; van Benthem, Koen J. ; Magnusson, Arni ; Berg, Casper Willestofte ; Nielsen, Anders ; Skaug, Hans J. ; Machler, Martin ; Bolker, Benjamin M. / glmmTMB balances speed and flexibility among packages for Zero-inflated Generalized Linear Mixed Modeling. In: The R Journal. 2017 ; Vol. 9, No. 2. pp. 378-400.
@article{77b9a508716047a8ab9bc8acfbaf0c48,
title = "glmmTMB balances speed and flexibility among packages for Zero-inflated Generalized Linear Mixed Modeling",
abstract = "Count data can be analyzed using generalized linear mixed models when observations are correlated in ways that require random effects. However, count data are often zero-inflated, containing more zeros than would be expected from the typical error distributions. We present a new package, glmmTMB, and compare it to other R packages that fit zero-inflated mixed models. The glmmTMB package fits many types of GLMMs and extensions, including models with continuously distributed responses, but here we focus on count responses. glmmTMB is faster than glmmADMB, MCMCglmm, and brms, and more flexible than INLA and mgcv for zero-inflated modeling. One unique feature of glmmTMB (among packages that fit zero-inflated mixed models) is its ability to estimate the Conway-Maxwell-Poisson distribution parameterized by the mean. Overall, its most appealing features for new users may be the combination of speed, flexibility, and its interface's similarity to lme4.",
keywords = "COMPUTER, STATISTICS, MAXWELL-POISSON DISTRIBUTION, COUNT DATA, BAYESIAN-INFERENCE, R PACKAGE, REGRESSION, ECOLOGY, EVOLUTION, ABUNDANCE",
author = "Brooks, {Mollie Elizabeth} and Kasper Kristensen and {van Benthem}, {Koen J.} and Arni Magnusson and Berg, {Casper Willestofte} and Anders Nielsen and Skaug, {Hans J.} and Martin Machler and Bolker, {Benjamin M.}",
year = "2017",
doi = "10.32614/RJ-2017-066",
language = "English",
volume = "9",
pages = "378--400",
journal = "The R Journal",
issn = "2073-4859",
publisher = "R Foundation for Statistical Computing",
number = "2",

}

glmmTMB balances speed and flexibility among packages for Zero-inflated Generalized Linear Mixed Modeling. / Brooks, Mollie Elizabeth; Kristensen, Kasper; van Benthem, Koen J.; Magnusson, Arni; Berg, Casper Willestofte; Nielsen, Anders; Skaug, Hans J.; Machler, Martin; Bolker, Benjamin M.

In: The R Journal, Vol. 9, No. 2, 2017, p. 378-400.

Research output: Contribution to journalJournal articleResearchpeer-review

TY - JOUR

T1 - glmmTMB balances speed and flexibility among packages for Zero-inflated Generalized Linear Mixed Modeling

AU - Brooks, Mollie Elizabeth

AU - Kristensen, Kasper

AU - van Benthem, Koen J.

AU - Magnusson, Arni

AU - Berg, Casper Willestofte

AU - Nielsen, Anders

AU - Skaug, Hans J.

AU - Machler, Martin

AU - Bolker, Benjamin M.

PY - 2017

Y1 - 2017

N2 - Count data can be analyzed using generalized linear mixed models when observations are correlated in ways that require random effects. However, count data are often zero-inflated, containing more zeros than would be expected from the typical error distributions. We present a new package, glmmTMB, and compare it to other R packages that fit zero-inflated mixed models. The glmmTMB package fits many types of GLMMs and extensions, including models with continuously distributed responses, but here we focus on count responses. glmmTMB is faster than glmmADMB, MCMCglmm, and brms, and more flexible than INLA and mgcv for zero-inflated modeling. One unique feature of glmmTMB (among packages that fit zero-inflated mixed models) is its ability to estimate the Conway-Maxwell-Poisson distribution parameterized by the mean. Overall, its most appealing features for new users may be the combination of speed, flexibility, and its interface's similarity to lme4.

AB - Count data can be analyzed using generalized linear mixed models when observations are correlated in ways that require random effects. However, count data are often zero-inflated, containing more zeros than would be expected from the typical error distributions. We present a new package, glmmTMB, and compare it to other R packages that fit zero-inflated mixed models. The glmmTMB package fits many types of GLMMs and extensions, including models with continuously distributed responses, but here we focus on count responses. glmmTMB is faster than glmmADMB, MCMCglmm, and brms, and more flexible than INLA and mgcv for zero-inflated modeling. One unique feature of glmmTMB (among packages that fit zero-inflated mixed models) is its ability to estimate the Conway-Maxwell-Poisson distribution parameterized by the mean. Overall, its most appealing features for new users may be the combination of speed, flexibility, and its interface's similarity to lme4.

KW - COMPUTER

KW - STATISTICS

KW - MAXWELL-POISSON DISTRIBUTION

KW - COUNT DATA

KW - BAYESIAN-INFERENCE

KW - R PACKAGE

KW - REGRESSION

KW - ECOLOGY

KW - EVOLUTION

KW - ABUNDANCE

U2 - 10.32614/RJ-2017-066

DO - 10.32614/RJ-2017-066

M3 - Journal article

VL - 9

SP - 378

EP - 400

JO - The R Journal

JF - The R Journal

SN - 2073-4859

IS - 2

ER -