Skip to contents

Simulate statnet Network Object

Source: R/simulate_network.R
simulate_network.Rd

Generate a statnet network object using common graph algorithms with realistic node names from human names or learning states.

Usage

simulate_network(
  n = NULL,
  model = c("er", "ba", "ws", "sbm", "reg", "grg", "ff"),
  name_source = c("human", "states"),
  regions = "all",
  categories = "all",
  names = NULL,
  directed = FALSE,
  weighted = FALSE,
  weights = c(0.1, 1),
  p = 0.1,
  m = NULL,
  power = 1,
  m_ba = 2,
  nei = 2,
  p_rewire = 0.05,
  blocks = 3,
  p_within = 0.3,
  p_between = 0.05,
  k = 4,
  radius = 0.25,
  fw = 0.35,
  bw = 0.32,
  seed = NULL
)

Arguments

n

Integer or NULL. Number of nodes. If NULL (default), randomly selects between 20-50 nodes.

model

Character. Graph generation algorithm:

  • "er": Erdos-Renyi random graph

  • "ba": Barabasi-Albert scale-free network

  • "ws": Watts-Strogatz small-world network

  • "sbm": Stochastic Block Model (community structure)

  • "reg": Regular graph (fixed degree)

  • "grg": Geometric Random Graph (spatial)

  • "ff": Forest Fire (growing network)

Default: "er".

name_source

Character. Source for node names:

  • "human": Culturally diverse human names from GLOBAL_NAMES

  • "states": Learning action verbs from LEARNING_STATES

Default: "human".

regions

Character vector. Regions to sample human names from (only used when name_source = "human"). Can be specific regions (e.g., "arab", "east_asia"), shortcuts (e.g., "europe", "africa", "asia"), or "all". See list_name_regions. Default: "all".

categories

Character vector. Learning state categories (only used when name_source = "states"). Options: "metacognitive", "cognitive", "behavioral", "social", "motivational", "affective", "group_regulation", "lms", or "all". Default: "all".

names

Character vector or NULL. Custom node names. Overrides name_source if provided. Default: NULL.

directed

Logical. If TRUE, generate directed network. Default: FALSE.

weighted

Logical. If TRUE, add random edge weights. Default: FALSE.

weights

Numeric vector of length 2. Weight range [min, max]. Default: c(0.1, 1.0).

p

Numeric. Edge probability for Erdos-Renyi model. Default: 0.1.

m

Integer or NULL. Fixed number of edges for Erdos-Renyi. Overrides p if provided. Default: NULL.

power

Numeric. Attachment power for Barabasi-Albert. Default: 1.

m_ba

Integer. Edges per new vertex for Barabasi-Albert. Default: 2.

nei

Integer. Neighborhood size for Watts-Strogatz. Default: 2.

p_rewire

Numeric. Rewiring probability for Watts-Strogatz. Default: 0.05.

blocks

Integer. Number of blocks for SBM. Default: 3.

p_within

Numeric. Within-block edge probability for SBM. Default: 0.3.

p_between

Numeric. Between-block edge probability for SBM. Default: 0.05.

k

Integer. Degree for regular graphs. Default: 4.

radius

Numeric. Connection radius for geometric random graph. Default: 0.25.

fw

Numeric. Forward burning probability for Forest Fire. Default: 0.35.

bw

Numeric. Backward burning factor for Forest Fire. Default: 0.32.

seed

Integer or NULL. Random seed for reproducibility. Default: NULL.

Value

A network object (class "network") with vertex names and optional edge weights.

Details

This function generates networks using igraph algorithms internally, then converts to statnet's network class. The resulting object is compatible with all sna and network package functions.

Vertex names are stored in the "vertex.names" attribute and can be accessed with network.vertex.names().

See also

simulate_igraph for igraph objects, simulate_matrix for adjacency matrices, simulate_edge_list for edge list data frames.

Examples

if (FALSE) { # \dontrun{
library(network)
library(sna)

# Default: Erdos-Renyi with human names
net <- simulate_network(n = 20, seed = 42)
class(net)  # "network"
network.vertex.names(net)  # Diverse human names

# Names from specific regions
net_asia <- simulate_network(n = 20, regions = "asia", seed = 42)
network.vertex.names(net_asia)  # Asian names

# Use learning state names instead
net_states <- simulate_network(n = 15, name_source = "states", seed = 42)
network.vertex.names(net_states)  # Action verbs

# Scale-free for SNA analysis
net_sf <- simulate_network(n = 50, model = "ba", seed = 42)
betweenness(net_sf)
closeness(net_sf)

# Community structure
net_sbm <- simulate_network(n = 30, model = "sbm", blocks = 3, seed = 42)

# Weighted network
net_w <- simulate_network(n = 20, model = "er", weighted = TRUE, seed = 42)
net_w %e% "weight"  # Edge weights
} # }