do_closest_neighbor_graph: Construct a new graph for a ClusterSet object (internal)

In dputhier/dbfmcl: The scigenex package (Single-Cell Informative GENe Explorer)

Construct a new graph for a ClusterSet object (internal)

Description

This function constructs a new graph for a given ClusterSet object based on selected genes. The graph is constructed using the Density K-Nearest Neighbor (DKNN) method.

Usage

do_closest_neighbor_graph(
  object = NULL,
  k = 5,
  output_path = NULL,
  name = NULL
)

Arguments

object	A ClusterSet object.
k	The number of nearest neighbors to consider for each gene.
output_path	a character indicating the path where the output files will be stored.
name	a character string giving the name for the output files. If NULL, a random name is generated.

Value

A ClusterSet object.

Examples

# Set verbosity to 1 to display info messages only.
set_verbosity(1)

# Create a matrix with 4 signatures
m <- create_4_rnd_clust()

# Select informative features
res <- select_genes(m[1:600,],
                    distance = "pearson",
                    k = 75,
                    noise_level = 0.3,
                    fdr = 1e-8,
                    row_sum = -Inf)
                    
# Construct a new graph based on genes selected with select_genes()
res <- do_closest_neighbor_graph(object = res, k = 5)

dputhier/dbfmcl documentation built on Dec. 20, 2024, 1:59 a.m.