run_optim_from_n: Run maximum likelihood esimation starting from the number of...

In TheoPannetier/DDvTDtools: Compare diversity-dependent (DD) and time-dependent (TD) diversification models

Description

An alternative version of run_optim() where the initial value of K is set to N, the nb of tips in the trees, in order to ease convergence to the maximum likelihood.

Usage

run_optim_from_n(
  sim,
  optim,
  para,
  rangemc = 1:1000,
  methode = "ode45",
  optimmethod = "subplex",
  tol = rep(1e-06, 3),
  jobID = NA,
  num_cycles = 1,
  save_results = TRUE,
  return_results = FALSE,
  cond = 1
)

Arguments

sim	character, name of the simulation model, either "DD" or "TD".
optim	character, name of the optimisation model, either "DD" or "TD".
para	numeric, a four-digits number coding for a set of four parameter values. Refer to arg_para() doc for details, and call arg_para() to see possible inputs.
rangemc	numeric vector, a set of tree indices ranging from 1 to 1000.
methode	likelihood solving methode, passed to DDD::dd_loglik() / DDD::bd_loglik().
optimmethod	optimisation algorithm, passed to DDD::dd_ML() / DDD::bd_ML().
tol	optimisation tolerance, passed to DDD::dd_ML() / DDD::bd_ML().
jobID	SLURM job ID passed when the function is called from a cluster script.
num_cycles	number of cycles of optimisation, passed to DDD::dd_ML() / DDD::bd_ML().
save_results	logical. Should the results be saved to outputfile?
return_results	logical. Should the results be returned?
cond	code specifying how the likelihood is conditioned. Passed to DDD::dd_loglik() / DDD::bd_loglik().

Author(s)

Théo Pannetier

TheoPannetier/DDvTDtools documentation built on Oct. 22, 2020, 2:31 p.m.