metricPerformance: Summarize metric performance of a series of summarized...

In eliotmiller/metricTester: Test Metric and Null Model Statistical Performance

Description

Summarizes metric performance after reading in and testing simulation results with either sesIndiv or plotOverall.

Usage

metricPerformance(summarized.results, nulls, concat.by = "both")

Arguments

summarized.results	The results of a call to sesIndiv() or plotOverall(). If from plotOverall, the results must include the standard three spatial simulations (random, filtering, competition), but must contain no other spatial simulations.
nulls	By default, this function summarizes metric performance over all tested null models. Alternatively, user can supply a vector of null model names to summarize the results over.
concat.by	Default is "both". Meaning that if both plot-level and richness-level summary statistics are available, then performance will be average over both types of results. Alternatively (and perhaps preferably), either "plot" or "richness" can be provided and, assuming that concatenation option was specified in the multiLinker runs, performance results will be summarized just over that specified option.

Details

If an overall picture of metric performance is desired, this function can provide it. It can also be used to summarize metric performance over a specific subset of simulations, null models, and concatenation options. If provided with the results of a call to plotOverall, the options are more limited. Currently, if provided with such a result, the assumption is that there are three spatial simulations, "random", "filtering", and "competition". It then assumes that any clustered or overdispersed plots for the random simulation, or any overdispersed or clustered for the filtering or competition simulations, respectively, count as typeI errors. It assumes that any plots that are not clustered or overdispersed for the filtering or competition simulations, respectively, count as typeII errors.

Value

A data frame of summarized results

References

Miller, E. T., D. R. Farine, and C. H. Trisos. 2016. Phylogenetic community structure metrics and null models: a review with new methods and software. Ecography DOI: 10.1111/ecog.02070

Examples

#not run
#results <- readIn()
#summ <- sesIndiv(results)
#examp <- metricPerformance(summ)

eliotmiller/metricTester documentation built on Dec. 16, 2019, 12:39 p.m.