centrality_stability: Centrality Stability Coefficient (CS-coefficient)

In Nestimate: Network Estimation, Bootstrap, and Higher-Order Analysis

Centrality Stability Coefficient (CS-coefficient)

Description

Estimates the stability of centrality indices under case-dropping. For each drop proportion, sequences are randomly removed and the network is re-estimated. The correlation between the original and subset centrality values is computed. The CS-coefficient is the maximum proportion of cases that can be dropped while maintaining a correlation above threshold in at least certainty of bootstrap samples.

For transition methods, uses pre-computed per-sequence count matrices for fast resampling. Strength centralities (InStrength, OutStrength) are computed directly from the matrix without igraph.

Usage

centrality_stability(
  x,
  measures = c("InStrength", "OutStrength", "Betweenness"),
  iter = 1000L,
  drop_prop = seq(0.1, 0.9, by = 0.1),
  threshold = 0.7,
  certainty = 0.95,
  method = "pearson",
  centrality_fn = NULL,
  loops = FALSE,
  normalize = FALSE,
  invert = TRUE,
  normalize_diffusion = TRUE,
  seed = NULL
)

Arguments

x	A netobject from build_network.
measures	Character vector. Centrality measures to assess. Defaults to c("InStrength", "OutStrength", "Betweenness"). Pass "all" for every built-in measure: "OutStrength", "InStrength", "ClosenessIn", "ClosenessOut", "Closeness", "Betweenness", "BetweennessRSP", "Diffusion", and "Clustering". The legacy aliases "InCloseness" and "OutCloseness" are also accepted. Custom measures beyond these are valid only when a centrality_fn is supplied to resolve them.
iter	Integer. Number of bootstrap iterations per drop proportion (default: 1000).
drop_prop	Numeric vector. Proportions of cases to drop (default: seq(0.1, 0.9, by = 0.1)).
threshold	Numeric. Minimum correlation to consider stable (default: 0.7).
certainty	Numeric. Required proportion of iterations above threshold (default: 0.95).
method	Character. Correlation method: "pearson", "spearman", or "kendall" (default: "pearson").
centrality_fn	Optional function. A custom centrality function that takes a weight matrix and returns a named list of centrality vectors. When NULL (default), all built-in measures are computed internally: "InStrength"/"OutStrength" via colSums/rowSums, and "Betweenness"/ "InCloseness"/"OutCloseness"/"Closeness" via an internal Floyd-Warshall shortest-path routine. When provided, the function is called as centrality_fn(mat) and is used only for requested measures that are not one of the six built-ins; it should return a named list (e.g., list(my_metric = ...)).
loops	Logical. If FALSE (default), self-loops (diagonal) are excluded from centrality computation. This does not modify the stored matrix.
normalize	Logical. Range-normalize all requested measures using the same transformation as tna::centralities(normalize = TRUE). Default: FALSE.
invert	Logical. Invert weights for shortest-path measures? Default: TRUE, matching tna.
normalize_diffusion	Logical. Range-normalize Diffusion even when normalize = FALSE. Default: TRUE.
seed	Integer or NULL. RNG seed for reproducibility.

Value

An object of class "net_stability" containing:

cs

Named numeric vector of CS-coefficients per measure.

correlations

Named list of matrices (iter x n_prop) of correlation values per measure.

measures

Character vector of measures assessed.

drop_prop

Drop proportions used.

threshold

Stability threshold.

certainty

Required certainty level.

iter

Number of iterations.

method

Correlation method.

See Also

build_network, network_reliability

Examples

net <- build_network(data.frame(V1 = c("A","B","C","A"),
  V2 = c("B","C","A","B")), method = "relative")
cs <- centrality_stability(net, iter = 10, drop_prop = 0.3)

seqs <- data.frame(
  V1 = sample(LETTERS[1:4], 30, TRUE), V2 = sample(LETTERS[1:4], 30, TRUE),
  V3 = sample(LETTERS[1:4], 30, TRUE), V4 = sample(LETTERS[1:4], 30, TRUE)
)
net <- build_network(seqs, method = "relative")
cs <- centrality_stability(net, iter = 100, seed = 42,
  measures = c("InStrength", "OutStrength"))
print(cs)

Nestimate documentation built on July 11, 2026, 1:09 a.m.