Learn R Programming

TMB (version 1.9.25)

Template Model Builder: A General Random Effect Tool Inspired by 'ADMB'

Description

With this tool, a user should be able to quickly implement complex random effect models through simple C++ templates. The package combines 'CppAD' (C++ automatic differentiation), 'Eigen' (templated matrix-vector library) and 'CHOLMOD' (sparse matrix routines available from R) to obtain an efficient implementation of the applied Laplace approximation with exact derivatives. Key features are: Automatic sparseness detection, parallelism through 'BLAS' and parallel user templates.

Copy Link

Version

Install

install.packages('TMB')

Monthly Downloads

56,090

Version

1.9.25

License

GPL-2

Issues

70

Pull Requests

5

Stars

195

Forks

90

Maintainer

Kasper Kristensen

Last Published

August 22nd, 2026

Functions in TMB (1.9.25)

newtonOption

Set newton options for a model object.
newton

Generalized newton optimizer.
gdbsource

Source R-script through gdb to get backtrace.
openmp

Control number of OpenMP threads used by a TMB model.
plot.tmbprofile

Plot likelihood profile.
config

Get or set internal configuration variables
summary.checkConsistency

Summarize output from checkConsistency
confint.tmbprofile

Profile based confidence intervals.
precompile

Precompile the TMB library in order to speed up compilation of templates.
print.sdreport

Print brief model summary
print.checkConsistency

Print output from checkConsistency
tmbroot

Compute likelihood profile confidence intervals of a TMB object by root-finding
summary.sdreport

summary tables of model parameters
runExample

Run one of the test examples.
runSymbolicAnalysis

Run symbolic analysis on sparse Hessian
sdreport

General sdreport function.
tmbprofile

Adaptive likelihood profiling.
template

Create cpp template to get started.
SR

Sequential reduction configuration
checkConsistency

Check consistency and Laplace accuracy
TMB.Version

Version information on API and ABI.
FreeADFun

Free memory allocated on the C++ side by MakeADFun.
GK

Gauss Kronrod configuration
compile

Compile a C++ template to DLL suitable for MakeADFun.
benchmark

Benchmark parallel templates
as.list.sdreport

Convert estimates to original list format.
MakeADFun

Construct objective functions with derivatives based on a compiled C++ template.
Rinterface

Create minimal R-code corresponding to a cpp template.
normalize

Normalize process likelihood using the Laplace approximation.
dynlib

Add dynlib extension
oneStepPredict

Calculate one-step-ahead (OSA) residuals for a latent variable model.