Queue projects for marginal likelihood estimation

EnqueueML() mirrors the workflow of EnqueueMC() but instead enqueues jobs for marginal likelihood estimation using the stepping-stone or path-sampling methods in RevBayes.

EnqueueML(projects, models, overwrite = TRUE)

Arguments

projects

Character vector of project identifiers.

models

Character vector of RevBayes model identifiers.

overwrite

Logical. Whether to overwrite existing processed matrices. Defaults to TRUE.

Value

EnqueueML() invisibly returns NULL. The function is called for its side-effect of preparing and queuing marginal-likelihood analyses on the configured remote cluster.

Details

Each project’s matrix is prepared with PrepareMatrix() and paired with model scripts generated via RevBayes(). If the marginal likelihood output (StoneFile()) is absent, SLURM job files are created using MakeSlurm(ml = TRUE).

This function requires that SSH credentials and GitHub authentication have been configured in .Renviron, as described in MakeRepo().

See also

EnqueueMC(), RevBayes(), MakeSlurm(), PrepareMatrix()