Skip to contents

Run diffusion scoring

diffuse_scores.Rd

Diffuses input scores over the precomputed kernel. Requires build_kernel() to have been called first.

Usage

diffuse_scores(
  object,
  input,
  background = NULL,
  method = "raw",
  n_perm = 1000L,
  seed = 42L
)

Arguments

object

A DiffusionScores object with a built kernel.

input

Named numeric vector or named matrix. Names must correspond to node names in the graph. For a vector, treated as a single input column. For a matrix, rows are background nodes and columns are separate input sets.

background

Character vector. Node names forming the background set. Defaults to all nodes in input.

method

Character. One of "raw", "z", or "mc". Defaults to "raw".

n_perm

Integer. Number of permutations for method = "mc". Defaults to 1000L.

seed

Integer. RNG seed for method = "mc". Defaults to 42L.

Value

The DiffusionScores object with scores populated as a data.table with columns node and one score column per input.