bundles {multiplex}

Bundle Class Patterns

Description

Classify the Bundle class patterns in a system of multiple relations

Usage

bundles(x, loops = FALSE, smpl = FALSE, lb2lb = TRUE, 
        collapse = FALSE, sep)

Arguments

x

an array; usually with three dimensions of stacked matrices where the multiple relations are placed.

loops

(logical) whether or not the loops should be considered as a particular bundle

smpl

(logical) simplify the strings of relations? Default no.

lb2lb

(logical) should the labels of the nodes be included in the output? (default yes).

collapse

(logical) collapse the distinct levels of relations in the network? (default no).

sep

(optional) the pair separator for the pairwise relations

Details

A bundle is a particular type of pattern made of relations at different levels that is binding a pair of nodes or actors in a network of relationships. A bundle class is a dyadic configuration resulting from the mixture of the direction and the types of ties between the nodes or actors. There are in total seven dyadic configuration classes, which are null, asymmetric, reciprocal, tie entrainment, tie exchange, mixed, and the full bundle pattern. This function provides detailed information about the bundle class patterns in multiple networks as lists of pair relations among the nodes or actors, except for the ‘null’ pattern.

In case that the nodes are not labeled, then an identification number will be assigned according to the nodes' location in the array representation and as well when the lb2lb option is set to FALSE. This function assumes that the network is directed, and self ties can also be considered in the output. Long string labels can be simplified with smpl, whereas the collapse option blurs the levels in the strings.

Value

An object of 'Rel.Bundles' class with the distinct bundle class patterns.

asym

asymmetric ties

recp

reciprocal ties

tent

tie entrainment

txch

tie exchange

mixed

mixed

full

full

loops

loops (if chosen)

Note

The input array for this function is always dichotomized, and it is possible to obtain the total number of occurrences in each bundle class pattern with the bundle.census function.

Author(s)

Antonio Rivero Ostoic

References

Ostoic, J. A. R. ‘Dyadic Patterns in Multiple Networks,’ Advances in Social Networks Analysis and Mining, International Conference on, 475-481. 2011.

See Also

bundle.census, summaryBundles, transf.

Examples

## Create the data: two binary relations among three elements
arr <- round( replace( array(runif(18), c(3,3,2)), array(runif(18),
       c(3,3,2))>.8, 3 ) )

## Establish the different bundles
bundles(arr)