as_mixing_matrix: Construct a graph object's mixing matrix.

In knapply/homophily: Measuring Network Homophily

Description

Using a categorical vertex attribute, construct a matrix depicting network mixing patterns.

Usage

as_mixing_matrix(g, dim1, dim2 = NULL, direction = c("all", "out",
  "in"))

Arguments

g	<igraph>, <network>
dim1	Either the name of a node/vertex attribute or vector of same length as the number of nodes in g.
dim2	If not provided, the same as dim1. Otherwise, a vector of the same length as the number of nodes in g.
direction	<chr> The direction of ties to take into a account. "all", "out", or "in"

Details

• If g is undirected, dim1 corresponds to the rows of the returned mixing matrix and dim2 corresponds to the columns.

  • If dim1 and dim2 refer to the same dimension/attribute, only the upper triangle of the returned mixing matrix is used.

• If g is directed, the rows of the returned mixing matrix correspond to the source of ties while the columns correspond to the target of ties.

Value

• <dtrMatrix>

  • g is undirected and dim1 and dim2 refer to same dimension.

• <dgeMatrix>

  • g is directed or dim1 and dim2 refer to different dimensions.

• <ddiMatrix>

  • g exhibits 100

Author(s)

Brendan Knapp [e-mail]

Examples

as_mixing_matrix(jemmah_islamiyah, dim1 = "role")
as_mixing_matrix(jemmah_islamiyah, dim1 = "name", dim2 = "role")

data("sampson", package = "ergm")
as_mixing_matrix(samplike, dim1 = "vertex.names", dim2 = "group",
                 direction = "all")
as_mixing_matrix(samplike, dim1 = "vertex.names", dim2 = "group",
                 direction = "out")
as_mixing_matrix(samplike, dim1 = "vertex.names", dim2 = "group", 
                 direction = "in")

knapply/homophily documentation built on Nov. 4, 2019, 3:54 p.m.