metricTester (version 1.0.0)

varyX: Calculate alpha or beta metrics across a set of parameters

Description

Takes a specified set of parameters, and holds those constant while varying one of the parameters to see variance in community structure metrics.

Usage

varyX(alpha = TRUE, tree.size, richness, delta, abundances, beta.iterations,
  iterations, cores)

Arguments

alpha

Whether to calculate alpha or beta phylogenetic community structure metrics. Default is TRUE. Set to FALSE for beta metrics.

tree.size

Either a single number, or if the aim is to vary tree size while holding other parameters constant, then a list of numbers of species desired in each total tree. If provided as a vector of numbers, will be coerced to a list.

richness

Either the number of species to be placed in each plot, or if the aim is to vary richness while holding other parameters constant, then a list of vectors of number of species to be placed in each plot. See examples.

delta

Either a value for the delta transformation (Pagel 1999), or if the aim is to vary tree shape while holding other parameters constant, then a list of numbers of delta transformations. If provided as a vector of numbers, will be coerced to a list. Values g

abundances

Either a vector of abundances, or if the aim is to vary the abundance distribution function, then a list of vectors of abundances. See examples.

beta.iterations

Because the type of beta-level phylogenetic community structure metrics used here return a single value per community data matrix, it is not possible to look for inter-metric correlations with only a single matrix and tree. To deal with this, the same tre

iterations

How many times to simulate the given set of parameters. For instance, with a single tree size, richness.vector, delta, and two sets of abundances, and 10 iterations, the result will be a list with 10 elements. Each of those 10 elements will be a list of t

cores

Number of cores to be used for parallel processing. The iteration aspect of the function is parallelized, so for efficiency purposes, it is best to run this over numerous iterations rather than repeating the same parameter numerous times (e.g., rather tha

Value

• A list of lists of data frames.

Details

If given a small value, e.g. 0.1, the delta parameter (tree shape) can occasionally result in oddly formatted trees that would cause errors. To deal with this, there is an internal check that will recreate a new tree and re-scale it with the desired delta. This has not been tested at delta < 0.1, and is currently programmed with a while loop. Care should be taken not to get R stuck in an indefinite loop at delta values even lower than 0.1

References

Miller, E. T., D. R. Farine, and C. H. Trisos. 2015. Phylogenetic community structure metrics and null models: a review with new methods and software. bioRxiv 025726.

Examples

#example of how to vary tree size
system.time(temp <- varyX(alpha=TRUE, tree.size=c(59, 100),
	richness=40:59, delta=1,
	abundances=round(rlnorm(5000, meanlog=2, sdlog=1)) + 1, iterations=2, cores=1))

#example of how to vary richness
#not run
#system.time(temp <- varyX(alpha=TRUE, tree.size=59,
	#richness=list(30:39, 40:49), delta=1,
	#abundances=round(rlnorm(5000, meanlog=2, sdlog=1)) + 1, iterations=2, cores=1))

#example of how to vary tree shape
#not run
#system.time(temp <- varyX(alpha=TRUE, tree.size=59,
	#richness=40:59, delta=c(0.1,10),
	#abundances=round(rlnorm(5000, meanlog=2, sdlog=1)) + 1, iterations=2, cores=1))

#example of how to vary abundance
#not run
#inputAbunds <- list(rep(2,5000), (round(rnorm(5000, 50, 15)) + 1),
	#(round(rlnorm(5000, meanlog=2, sdlog=1)) + 1))

#system.time(temp <- varyX(alpha=TRUE, tree.size=59,
	#richness=40:59, delta=1, abundances=inputAbunds, iterations=2, cores=1))