sequential.corr: Sequential correlation

In AMBarbosa/DeadCanMove: Assess How Spatial Roadkill Patterns Change with Temporal Sampling Scheme

Description

This function calculates the correlation between the hotspots obtained from each submatrix and those of the baseline (sub)matrix of the corresponding group.

Usage

sequential.corr(hotspots.list, hotspots.thresholds, 
comp.method = "Phi", baseline.interval = 1, baseline.gap = 0, 
messages = "TRUE")

Arguments

hotspots.list	a list of hotspot tables resulting from the sequential.hotspots function
hotspots.thresholds	a matrix of hotspots thresholds (element 2 of the results of the hotspot.numbers function)
comp.method	characer value indicating the correlation coefficient to use; type binary.comp.methods() for available options
baseline.interval	the sampling interval with which to correlate all the other sampling intervals for each group; defaults to 1 (take every sample)
baseline.gap	the sampling gap with which to correlate all other sampling schemes for each group; defaults to 0 (no gap between samples)
messages	logical, whether to display messages

Value

A matrix of correlations (or whatever index was defined in method) between the hotspots obtained for each group and sampling scheme, and the hotspots obtained from the baseline data for the group under analysis.

Note

This function currently works only for hotspots of submats created with 'sampl.interval', not 'window.size' and 'gap.size'. See submatrix, sequential.submatrix, and check that your names(hotspots.list[[1]]) are something like "group.intv1", not "group.w1.g2.s1".

Author(s)

A. Marcia Barbosa

See Also

binary.comparison

Examples

data(roadkills)

submats <- sequential.submatrix(dataset = roadkills,
sampl.columns = 4:ncol(roadkills), sampl.interval = 1:3,
group.column = "taxon", include.all.together = TRUE, 
remove.zeros = TRUE, keep.nonsampl.columns = TRUE, 
n.subsampl.columns = 120)

hsl <- sequential.hotspots(dataset = roadkills, submats = submats,
region.column = "segment", first.subsampl.col = 4, confidence = 0.95)

hsn <- hotspot.numbers(hotspots.list = hsl, sampl.intervals = 1:3,
groups = as.character(unique(roadkills$taxon)), 
include.all.together = TRUE, min.total.events = 0, 
min.hotspot.threshold = 2)

seqcorr <- sequential.corr(hotspots.list = hsl,
hotspots.thresholds = hsn$HS.threshold, comp.method = "Phi",
baseline.gap = 0, messages = "TRUE")

seqcorr

AMBarbosa/DeadCanMove documentation built on Nov. 3, 2021, 10:03 a.m.