This function implements methods for investigating the optimal number of
genetic clusters ('k') using the fast maximum-likelihood genetic clustering
approach described in Beugin et al (2018). The method runs
snapclust for varying values of 'k', and computes the requested
summary statistics for each clustering solution to assess goodness of
fit. The method is fully documented in a dedicated tutorial which can be
accessed using adegenetTutorial("snapclust").
snapclust.choose.k(max, ..., IC = AIC, IC.only = TRUE)An integer indicating the maximum number of clusters to seek;
snapclust will be run for all k from 2 to max.
Arguments passed to snapclust.
A function computing the information criterion for
snapclust objects. Available statistics are
AIC (default), AICc, and BIC.
A logical (TRUE by default) indicating if IC values only
should be returned; if FALSE, full snapclust objects are
returned.
Thibaut Jombart thibautjombart@gmail.com
The method is described in Beugin et al (2018) A fast likelihood
solution to the genetic clustering problem. Methods in Ecology and
Evolution tools:::Rd_expr_doi("10.1111/2041-210X.12968"). A dedicated
tutorial is available by typing adegenetTutorial("snapclust").
snapclust to generate individual clustering solutions,
and BIC.snapclust for computing BIC for snapclust
objects.