pou_simulation_study: Phylogenetic Simulation Study

In jpmeagher/sdsBAT:

Description

Type II Maximum Likelihood Estimation of hyperparameters for n_samples realisations of a Phylogenetic Ornstein-Uhlenbeck Process with given hyperparameters

Usage

pou_simulation_study(ln_hyperparameters, phylogenetic_tree, n_samples,
  logl_function = pou_logl_slow, optim_function = "optim",
  optim_method = "Nelder-Mead", lower_initialisation = c(0, 0, 0),
  upper_initialisation = c(1, 1, 1), n_restarts = 1)

Arguments

ln_hyperparameters	Vector of the natural logarithm of hyperparameters for an Ornstein-Uhlenbeck kernel in order log(phylogenetic noise, phylogenetic length-scale, non-phylogenetic noise)
phylogenetic_tree	Phylogenetic tree of relationships between observations
logl_function	function calculating the negative log likelihood of the observations to be minimised. Either 'pou_logl_slow' or 'pou_logl_fast'.
optim_function	String indicating which optimising function to use, 'optim' or 'uobyqa' from the minqa package.
optim_method	Method to be used with the optim optimiser.
lower_initialisation	Lower bound for the uniform random variable used to generate initial values.
upper_initialisation	Upper bound for the uniform random variable used to generate initial values.
n_restarts	Number of different initialisation values to test.

Value

A 4 x n_samples matrix. Hyperparameters are ordered in the rows as (phylogenetic noise, phylogenetic length-scale, non-phylogenetic noise). Each column is a separate sample.

jpmeagher/sdsBAT documentation built on May 5, 2019, 3:50 a.m.