tcor: Generates Correlated Traits

In ecolottery: Coalescent-Based Simulation of Ecological Communities

Description

Create two random vectors of traits correlated between each other or a vector of traits correlated to an existing one. The linear correlation is defined by the parameter rho.

Usage

tcor(n, rho = 0.5, mar.fun = rnorm, x = NULL, ...)

Arguments

n	the integer number of values to be generated.
rho	a numeric parameter defining the linear correlation between the two traits (default is 0.5). It must belong to the interval [-1, 1].
x	an vector of numeric values. Default is NULL.
mar.fun	a function defining the random generation for the trait distribution. Default is rnorm.
...	other arguments for the mar.fun() function.

Details

rho parameter is set to 0.5 by default. x = NULL by default. Code adapted from: http://stats.stackexchange.com/questions/15011/generate-a-random-variable-with-a-defined-correlation-to-an-existing-variable

Value

Return a data.frame with two numeric columns, each column defining a trait.

Author(s)

P. Denelle F. Munoz

Examples

# With no predefined trait
traits <- tcor(n = 10000, rho = 0.8)
plot(traits[, 1], traits[, 2])
cor(traits[, 1], traits[, 2])

# With existing trait
existing_trait <- rnorm(10000, 10, 1)
traits <- tcor(n = 10000, rho = 0.8, x = existing_trait)
plot(traits[, 1], traits[, 2])
cor(traits[, 1], traits[, 2])

Example output

[1] 0.795797
[1] 0.8021336

ecolottery documentation built on May 2, 2019, 9:34 a.m.