asnipe (version 1.1)

network_swap: Perform one (or more) random swap

Description

Performs one (or more) random swap on the data and re-calculates network, returning both the new network and the data stream

Usage

network_swap(association_data, data_format = "GBI", swaps = 1, 
	association_index = "SRI", association_matrix = NULL, locations = NULL, 
	classes = NULL, days = NULL, within_day = FALSE, within_location = FALSE, 
	within_class = FALSE)

Arguments

association_data

a K x N matrix of K groups (observations, gathering events, etc.) and N individuals (all individuals that are present in at least one group) OR a K x N x N array of sampling periods.

data_format

"GBI" expect a group by individual matrix, "SP" Expect a sampling periods array

swaps

number of swaps (default = 1000)

association_index

"SRI" Simple ratio index, "HWI" Half-weight index (more to come)

association_matrix

provide a starting association matrix (see details)

locations

K vector of locations defining the location of each group/event

classes

N vector of types or class of each individual (column) in the group by individual matrix (for subsetting)

days

K vector of day stamp for each event (can be integer or string representing any period of time)

within_day

if TRUE then permutations will be done within the time periods

within_location

if TRUE then permutations will be done within the given locations

within_class

if TRUE then permutations will be done within the given classes

Value

Returns a list containing an N x N matrix with the dyadic association rates of each pair of individuals after performing the swaps, and the N x N data stream post-swap, as two list elements.

Details

Performs one or more permutation swaps on the group by individual matrix as given by Whitehead (2008). In order to save on memory use, this function computers the number of swaps and returns the association matrix and the data stream resulting from these, thus not needing to create a large stack of networks to store each permutation. This can then be implemented in a loop as shown in the example below. Note that this method is quite a bit slower than the network_permutation function.

This implementation allows permutations (swaps) to be restricted to within any of three classes. Though each class is labelled, the function is flexible. Hence, days can represent any time period (months, hours, etc.). However, unlike the network_permutation, the subsetting of the data must be done outside of this function (for reasons that might be obvious) - see the example below.

References

Whitehead (2008) Analyzing Animal Societies

Examples

# load data
data("group_by_individual")
data("times")

# calculate network for data based on morning associations
network <- get_network(gbi, association_index="SRI", 
	times=times, start_time=0, end_time=max(times)/2)

# perform 100 permutations and calculate the coefficient
# of variance after each permutation.
# note that the subsetting is done outside of the function
library(raster)
cvs <- rep(NA,100)
network_perm = list(network,gbi[which(times <= max(times)/2),])
hours <- floor(times/3600)[which(times <= max(times)/2)]
for (i in 1:100) {
	network_perm <- network_swap(network_perm[[2]], swaps=1, 
		association_matrix=network_perm[[1]], days=hours, 
		within_day=TRUE)
	cvs[i] <- cv(network_perm[[1]])
}

# plot the results with the original network as a red dot
plot(cvs,pch=20,cex=0.5)
points(0,cv(network),cex=1,pch=20,col="red")