propagateLabelsDiffusion {sccore}

Estimate labeling distribution for each vertex, based on provided labels using a Random Walk on graph

Description

Estimate labeling distribution for each vertex, based on provided labels using a Random Walk on graph

Usage

propagateLabelsDiffusion(
  graph,
  labels,
  max.iters = 100,
  diffusion.fading = 10,
  diffusion.fading.const = 0.1,
  tol = 0.025,
  fixed.initial.labels = TRUE,
  verbose = TRUE
)

Arguments

graph

igraph graph object Graph input

labels

vector of factor or character labels, named by cell names

max.iters

integer Maximal number of iterations (default=100)

diffusion.fading

numeric Constant used for diffusion on the graph, exp(-diffusion.fading * (edge_length + diffusion.fading.const)) (default=10.0)

diffusion.fading.const

numeric Another constant used for diffusion on the graph, exp(-diffusion.fading * (edge_length + diffusion.fading.const)) (default=0.1)

tol

numeric Absolute tolerance as a stopping criteria (default=0.025)

fixed.initial.labels

boolean Prohibit changes of initial labels during diffusion (default=TRUE)

verbose

boolean Verbose mode (default=TRUE)

Value

matrix from input graph, with labels propagated

Examples

propagateLabelsDiffusion(conosGraph, labels=cellAnnotations)