find_communities: Community detection

In jmonlong/scCNAutils: Functions to analyze copy number aberrations in single-cell data

Community detection

Description

Build a KNN graph and run Louvain algorithm for community detection.

Usage

find_communities(
  pca_o,
  nb_pcs = 10,
  k = 100,
  gamma = 1,
  nreps = 1,
  nb_cores = 1
)

Arguments

pca_o	the output of run_pca
nb_pcs	the number of PCs to use. Default 10.
k	the number of nearest neighbor for the KNN graph. Default 100.
gamma	a vector of gamma. Default 1.
nreps	the number of repetition for each gamma, Default 1.
nb_cores	the number of processors to use. Default is 1.

Value

a list with:

comm	a data.frame with two columns: 'cell' and 'community'.
comm.all	a matrix with communities for each gamma
gamma	the list of input gamma corresponding to each comm.all column.
best.gamma	the gamma resulting on the highest ARI mean
ari.df	data.frame with ARI stats for each gamma

Author(s)

Jean Monlong

jmonlong/scCNAutils documentation built on May 3, 2022, 4:34 a.m.