Skip to contents

Convert igraph, network, or matrix objects to netify format

Source: R/to_netify.R
to_netify.Rd

Converts various network object types (igraph, network, matrices/arrays, or lists of these) into netify objects (also available as to_netify). Automatically extracts adjacency matrices and any nodal/dyadic attributes from the input objects.

Usage

to_netify(net_obj, weight = NULL, ...)

Arguments

net_obj

An R object to convert: igraph, network, matrix, array, or a list of these objects.

weight

Optional. Name of the weight attribute in net_obj to be used as the main edge weight in the netify object. Default is NULL. Important to specify for igraph and network objects as they do not have a default weight.

...

Additional arguments passed to new_netify. Can include nodal_data or dyad_data to override extracted attributes.

Value

A netify object with:

  • Adjacency matrix or list of matrices

  • Nodal attributes (if present in the input)

  • Dyadic attributes (if present in the input)

  • Weight specification (if provided)

Details

The function handles different input types:

  • igraph: Extracts adjacency matrix, vertex attributes as nodal data, and edge attributes as dyadic data

  • network: Extracts adjacency matrix, vertex attributes as nodal data, and edge attributes as dyadic data

  • matrix: Direct conversion, no attribute extraction

  • array: Assumes 3D arrays represent longitudinal networks

  • list: Must contain all objects of the same type (all igraph, all network, or all matrices)

For longitudinal data (lists or 3D arrays), the function creates a netify object with time-indexed components. Actor ordering is preserved from the input objects and made consistent across all components (adjacency, nodal, and dyadic data).

Note

When converting from igraph or network objects, specify the weight parameter to designate which edge attribute should be used as the primary edge weight in the netify object.

Author

Cassy Dorff, Shahryar Minhas

Examples

if (FALSE) { # \dontrun{
# From igraph
library(igraph)
g <- sample_gnp(10, 0.3)
E(g)$weight <- runif(ecount(g))
V(g)$type <- sample(c("A", "B"), vcount(g), replace = TRUE)

net <- to_netify(g, weight = "weight")

# From network  
library(network)
n <- network(rgraph(10, tprob = 0.3))
set.vertex.attribute(n, "group", sample(1:2, 10, replace = TRUE))

net <- to_netify(n)

# From matrix
adj_mat <- matrix(rnorm(100), 10, 10)
net <- to_netify(adj_mat)

# From list of matrices (longitudinal)
mat_list <- list(
  "2001" = matrix(rnorm(100), 10, 10),
  "2002" = matrix(rnorm(100), 10, 10)
)
net <- to_netify(mat_list)
} # }