From 7138fa41a5a459bcb44670b210ff4afb0e501d6c Mon Sep 17 00:00:00 2001 From: francois <francois.rousset@umontpellier.fr> Date: Tue, 5 Mar 2024 02:32:47 +0000 Subject: [PATCH] Update README.md --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 44be50a..73287b4 100644 --- a/README.md +++ b/README.md @@ -39,7 +39,7 @@ Also available here is the [Supplementary Appendix G](http://kimura.univ-montp2. For some substantial use of various features of `spaMM`, see e.g. the [IsoriX project](https://github.com/courtiol/IsoriX), or a story about [social dominance in hyaenas](https://doi.org/10.1038/s41559-018-0718-9), or [yet another depressing story about climate change](https://doi.org/10.1038/s41467-019-10924-4), or [the life-history of mothers of twins](https://doi.org/10.1038/s41467-022-30366-9), or a comparison of prediction by LMMs and by random-forest methods (in supplementary material of [a paper on protected area personnel](https://doi.org/10.1038/s41893-022-00970-0)), or [analyses of dyadic interactions in mandrills](https://doi.org/10.7554/eLife.79417). -Initial development drew inspiration from work by Lee and Nelder on _h_-likelihood and more elaborate approximations of likelihood (e.g. [Lee, Nelder & Pawitan](https://doi.org/10.1201/9781420011340), 2006; [Lee & Lee](http://dx.doi.org/10.1007/s11222-011-9265-9) 2012; see also [Molas and Lesaffre](http://dx.doi.org/10.1002/sim.3852), 2010). The latter two references, and `spaMM` itself, may actually be more widely understood as applications of Laplace approximations of likelihood than as applications of a distinctive _h_-likelihood concept. `spaMM` retains from Lee & Nelder's work several distinctive features, such as a concept of restricted likelihood applicable beyond LMMs, specific methods to fit models with non-gaussian random effects, structured dispersion models with random effects, and implementation of several variants of Laplace and PQL approximations. However, it has departed from them in various ways. Notably, the default likelihood and restricted likelihood approximations now go beyond those discussed in these works. For ML fits, it is the same Laplace approximation as in `TMB` ([Kristensen et al., 2016](https://doi.org/10.18637/jss.v070.i05)) and packages based on `TMB`, because `TMB` and `spaMM` (with default arguments) use the observed Hessian matrix of ``joint likelihood'' where the _h_-likelihood literature only considers the expected Hessian matrix. This makes a difference for GLM families with non-canonical link, or for response families not of the GLM class. +Initial development drew inspiration from work by Lee and Nelder on _h_-likelihood and more elaborate approximations of likelihood (e.g. [Lee, Nelder & Pawitan](https://doi.org/10.1201/9781420011340), 2006; [Lee & Lee](http://dx.doi.org/10.1007/s11222-011-9265-9) 2012; see also [Molas and Lesaffre](http://dx.doi.org/10.1002/sim.3852), 2010). The latter two references, and `spaMM` itself, may actually be more widely understood as applications of Laplace approximations of likelihood than as applications of a distinctive _h_-likelihood concept. `spaMM` retains from Lee & Nelder's work several distinctive features, such as a concept of restricted likelihood applicable beyond LMMs, specific methods to fit models with non-gaussian random effects, structured dispersion models with random effects, and implementation of several variants of Laplace and PQL approximations. However, it has departed from their work in various ways. Notably, the default likelihood and restricted likelihood approximations now go beyond those discussed in these works. For ML fits, it is the same Laplace approximation as in `TMB` ([Kristensen et al., 2016](https://doi.org/10.18637/jss.v070.i05)) and packages based on `TMB`, because `TMB` and `spaMM` (with default arguments) use the observed Hessian matrix of ``joint likelihood'' where the _h_-likelihood literature only considers the expected Hessian matrix. This makes a difference for GLM families with non-canonical link, or for response families not of the GLM class. ## Credits Initial development was supported by a PEPS grant from the CNRS and University of Montpellier. -- GitLab