Dynamic modeling of all kinds in R. These include models of processes in discrete time or continuous time. They also include processes that are linear or nonlinear. Latent variables can be continuous (e.g. state space models) or discrete (e.g. regime-switching models). The general approach involves maximum likelihood estimation of single- and multi-subject models of latent time series with the extended Kalman filter and Kim filter. The user provides recipes and data which are combined into a model that is then cooked to obtain free parameter estimates.
|Depends:||R (≥ 3.0.0), methods, ggplot2|
|Imports:||MASS, Matrix, numDeriv, xtable, latex2exp, grid, reshape2, plyr|
|Suggests:||testthat, roxygen2 (≥ 3.1)|
|Author:||Lu Ou [aut, cre], Michael D. Hunter [aut], Sy-Miin Chow [aut]|
|Maintainer:||Lu Ou <lzo114 at psu.edu>|
|License:||Apache License (== 2.0)|
|CRAN checks:||dynr results|
|Windows binaries:||r-devel: dynr_0.1.9-20.zip, r-release: dynr_0.1.9-20.zip, r-oldrel: dynr_0.1.9-20.zip|
|OS X Mavericks binaries:||r-release: dynr_0.1.9-20.tgz, r-oldrel: dynr_0.1.9-20.tgz|
|Old sources:||dynr archive|
Please use the canonical form https://CRAN.R-project.org/package=dynr to link to this page.