Wrapper function to generate label propagation parameters

Usage

params_label_propagation(
  alpha = 0.9,
  iter = 100L,
  tolerance = 1e-06,
  symmetrise = TRUE,
  symmetry_strategy = c("average", "avg", "min", "max"),
  max_hops = NULL
)

Arguments

alpha: Numeric. Controls the spreading strength. Higher values anchor labelled nodes more strongly to their original label. Defaults to 0.9.
iter: Integer. Maximum number of iterations to run. Defaults to 100L.
tolerance: Numeric. Convergence threshold. Stops early if the maximum change across all nodes falls below this value. Defaults to 1e-06.
symmetrise: Boolean. Whether to symmetrise the graph. Defaults to TRUE.
symmetry_strategy: String. Strategy to resolve weight conflicts when symmetrising. Only relevant when symmetrise = TRUE and edge weights are provided. One of c("average", "avg", "min", "max"). Defaults to "average".
max_hops: Integer or NULL. If provided, restricts label spreading to nodes within this many hops of any labelled node. Nodes beyond this limit remain all-zeroes. Defaults to NULL.

Value

A named list with the following elements:

alpha - Numeric. Controls the spreading strength. Higher values anchor labelled nodes more strongly to their original label. Defaults to 0.9.
iter - Integer. Maximum number of iterations to run. Defaults to 100L.
tolerance - Numeric. Convergence threshold. Stops early if the maximum change across all nodes falls below this value. Defaults to 1e-06.
symmetrise - Boolean. Whether to symmetrise the graph. Defaults to TRUE.
symmetry_strategy - String. Strategy to resolve weight conflicts when symmetrising. Only relevant when symmetrise = TRUE and edge weights are provided. One of c("average", "avg", "min", "max"). Defaults to "average".
max_hops - Integer or NULL. If provided, restricts label spreading to nodes within this many hops of any labelled node. Nodes beyond this limit remain all-zeroes. Defaults to NULL.