sim.pop: Simulate a population matrix

In dosreislab/mcmc3r: Tools to work with MCMCtree

Simulate a population matrix

Description

Simulate a population sample and return a list with (i) a matrix of size s x n, s specimens and n characters, (ii) a vector with the estimated population variances for each character, and (iii) the estimated shrinkage correlation matrix if the true correlation matrix is provided.

Usage

sim.pop(psample, n, c, R)

Arguments

psample	Numeric, number of specimens the simulated population sample should include.
n	Numeric, number of morphological traits to be simulated.
c	Numeric, vector with the variances for the species within a population (see details).
R	(Optional) matrix, correlation matrix. (see details).

Details

The parameter c is the population noise and it is used to sample n characters for each of the psample specimens from a normal distribution x ~ N(0,c). If the population noise is assumed to be the same for all the characters within the species, then the length of c is 1 and equals to the value of this variance. If it differs, then a vector of length n has to be provided specifying the variance for each of the characters

If a correlation matrix, R, is provided, then it is added to the population matrix. Note that the correlation matrix needs to be of class "matrix" and symmetric. You can take a look at data-raw/sim.R.R to follow the commands used to generate this matrix, object R.sim, which is used in the examples.

Value

$P	Matrix with the simulated population sample
$var	Vector with the estimated variances
$Rsh	Estimated shrinkage correlation matrix, only returned if R is provided

Author(s)

Sandra Alvarez-Carretero and Mario dos Reis

See Also

sim.morpho, write.morpho

Examples

# A) Simulation setup: Simulate a population with
#    psample = 20 specimens, and sample n = 100 characters with
#    a low population noise, c = 0.25.

     sim.pop( psample = 20, n = 100, c = 0.25 )

# B) Simulation setup: Simulate a population with
#    psample = 20 specimens, and sample n = 100 characters with
#    a low population noise, c = 0.25, and a low trait correlation
#    rho = 0.50 (correlation matrix that follows
#    the constant correlation model, i.e. all non-diagonal values
#    equal to rho).

     sim.pop( psample = 20, n = 100, c = 0.25, R = sim.R )

dosreislab/mcmc3r documentation built on March 29, 2024, 6:45 p.m.