Inference based on mixed-effect models, including generalized linear mixed models with spatial correlations and models with non-Gaussian random effects (e.g., Beta). Variation in residual variance (heteroscedasticity) can itself be represented by a generalized linear mixed model. Various approximations of likelihood or restricted likelihood are implemented, in particular h-likelihood (Lee and Nelder 2001 <doi:10.1093/biomet/88.4.987>) and Laplace approximation.
| Version: |
2.6.1 |
| Depends: |
R (≥ 3.2.0) |
| Imports: |
methods, stats, graphics, Matrix, MASS, proxy, Rcpp (≥
0.12.10), nlme, nloptr, pbapply |
| LinkingTo: |
Rcpp, RcppEigen |
| Suggests: |
maps, testthat, lme4, rsae, rcdd, pedigreemm, minqa, lpSolveAPI (≥ 5.5.0.14), foreach, multilevel, Infusion (≥
1.3.0), IsoriX (≥ 0.8.1), blackbox (≥ 1.1.25), gmp |
| Published: |
2019-01-14 |
| Author: |
François Rousset  [aut, cre,
cph],
Jean-Baptiste Ferdy [aut, cph],
Alexandre Courtiol [aut],
GSL authors [ctb] (src/gsl_bessel.*) |
| Maintainer: |
François Rousset <francois.rousset at umontpellier.fr> |
| License: |
CeCILL-2 |
| URL: |
https://www.r-project.org,
http://kimura.univ-montp2.fr/~rousset/spaMM.htm |
| NeedsCompilation: |
yes |
| Citation: |
spaMM citation info |
| Materials: |
NEWS |
| In views: |
Spatial |
| CRAN checks: |
spaMM results |