NEWS.md

In jSDM: Joint Species Distribution Models

jSDM 0.2.2

• New function jSDM_gaussian for fitting joint species distribution models from continuous Gaussian data, including an overdispersion parameter.
• Use of the R package terra instead of raster and sp, which are depreciated, in the vignette Estimation of Madagascar's plant biodiversity.

jSDM 0.2.1

• Add the possibility to consider only significant correlations in the get_residual_cor and plot_residual_cor functions.
• Documentation correction

jSDM 0.2.0

• New function jSDM_binomial_probit_sp_constrained which aims to improve the convergence of latent variable models fitting by selecting the species constrained to have positive values of factor loadings $\lambda$ and new vignette Bernoulli probit regression with selected constrained species to illustrate its use.
• New vignette Estimation of Madagascar's plant biodiversity available.

jSDM 0.1.2

• Add the possibility of considering an additional hierarchical level in the Bayesian models of the jSDM_binomial_probit, jSDM_binomial_logit and jSDM_poisson_log functions to take into account interactions between species-specific traits and the environment in estimating species effects.
• New vignette Bernoulli probit regression including species traits available.
• Separate the drawing of species effects beta and factor loading lambda in the functions jSDM_binomial_probit_block and jSDM_binomial_probit_block_long_format renamed jSDM_binomial_probit and jSDM_binomial_probit_long_format.
• New function plot_associations to plot species-species associations.

jSDM 0.1.1

• Use of roxygen2 for documentation and NAMESPACE
• Rename jSDM_binomial_probit_block the function jSDM_probit_block.
• New function jSDM_binomial_probit_block_long_format for fitting joint species distribution models from presence-absence data in long format able to handle missing observations, multiple visits at sites and to integer species traits as explanatory variables.
• New function jSDM_poisson_log for fitting joint species distribution models from abundance data inspired by Hui and Francis K. C. 2016 Methods in Ecology and Evolution (doi:10.1111/2041-210X.12514).
• New function jSDM_binomial_logit for fitting joint species distribution models from presence-absence data at multiple-visited sites using a bayesian inference method inspired by Albert, James H. and Chib Siddhartha 1993 Journal of the American Statistical Association (doi:10.1080/01621459.1993.10476321).
• Functions to fit models in which site effects are included as fixed effects, as random effects or not included and with or without latent variables.
• New function get_enviro_cor to extract covariances and correlations due to shared environmental responses.
• Complete jSDM-package documentation
• Add datasets (mosquitos, fungi, eucalypts, birds, mites, aravo).
• Seven new vignettes (Bayesian inference methods, Poisson log-linear regression, Bernoulli probit regression, Bernoulli probit regression with missing data and species traits, Binomial logistic regression, Running jSDM in parallel, Comparing SDMs and JSDMs and Comparison jSDM-Hmsc) are available.
• Complete and correct the vignette Comparison jSDM-boral.
• Add Code of conduct and Contributing section
• New package website available on GitHub: https://ecology.ghislainv.fr/jSDM/.

jSDM 0.1.0

• First version of the jSDM R package
• Use of Rcpp and C++ code for Gibbs sampling
• Use of GSL (RcppGSL) for random draws
• Use of Armadillo (RcppArmadillo) for vector and matrix operations
• Functions to fit models from Warton et al. 2014 Trends in Ecology and Evolution (doi:10.1016/j.tree.2015.09.007).
• We use pkgdown to build package website.
• Package website available on GitHub: https://ecology.ghislainv.fr/jSDM/.

jSDM documentation built on July 26, 2023, 5:21 p.m.