metricTester (version 1.2.2)

betaLinker: Run spatial simulations, null and beta metric calculations

Description

This function wraps a number of wrapper functions into one big metric + null tester function. Only a single test is performed, with results saved into memory.

Usage

betaLinker(no.taxa, arena.length, mean.log.individuals, length.parameter, sd.parameter, max.distance, proportion.killed, competition.iterations, no.plots, plot.length, randomizations, cores, simulations, nulls, metrics)

Arguments

no.taxa

The desired number of species in the input phylogeny

arena.length

A numeric, specifying the length of a single side of the arena

mean.log.individuals

Mean log of abundance vector from which species abundances will be drawn

length.parameter

Length of vector from which species' locations are drawn. Large values of this parameter dramatically decrease the speed of the function but result in nicer looking communities

sd.parameter

Standard deviation of vector from which species' locations are drawn

max.distance

The geographic distance within which neighboring indivduals should be considered to influence the individual in question

proportion.killed

The percent of individuals in the total arena that should be considered (as a proportion, e.g. 0.5 = half)

competition.iterations

Number of generations over which to run competition simulations

no.plots

Number of plots to place

plot.length

Length of one side of desired plot

randomizations

The number of randomized CDMs, per null, to generate. These are used to compare the significance of the observed metric scores.

cores

The number of cores to be used for parallel processing.

simulations

Optional list of named spatial simulation functions to use. These must be defined in the defineSimulations function. If invoked, this option will likely be used to run a subset of the defined spatial simulations.

nulls

Optional list of named null model functions to use. If invoked, this option will likely be used to run a subset of the defined null models.

metrics

Optional list of named metric functions to use. If invoked, this option will likely be used to run a subset of the defined metrics.

Value

A list with two elements. The first is a list of data frames, with one for each spatial simulation. These provide the observed beta metric scores for each spatial simulation. The second level is a list of lists, one for each spatial simulation. Each of these is a list of data frames. There is one data frame per null model, and it summarizes the randomized metric scores for that null model for that spatial simulation. Note that this is slightly different than the regular linker() function, which does not output these raw metric scores (that function calculates SES and CI as outputs).

Details

This function wraps a number of other wrapper functions into one big beta metric + null performance tester function. Only a single test is run, with results saved into memory. To perform multiple complete tests, use the multiLinker function, which saves results to file.

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

#below not run for timing issues on CRAN
#system.time(test <- betaLinker(no.taxa=50, arena.length=300, mean.log.individuals=2, 
	#length.parameter=5000, sd.parameter=50, max.distance=30, proportion.killed=0.2, 
#competition.iterations=3, no.plots=15, plot.length=30,
#randomizations=3, cores="seq",
#nulls=list("richness"=metricTester:::my_richnessNull,
#"frequency"=metricTester:::my_frequency)))