ClusterX: Fast clustering by automatic search and find of density peaks

In JinmiaoChenLab/cytofkit2: cytofkit2: an integrated mass cytometry data analysis pipeline. Cytofkit provides functions for data pre-processing, fast clustering, data visualization through linear or non-linear dimensionality reduction, automatic identification of cell subsets, and inference of the relatedness between cell subsets

Fast clustering by automatic search and find of density peaks

Description

This package implements the clustering algorithm described by Alex Rodriguez and Alessandro Laio (2014) with improvements of automatic peak detection and parallel implementation

Usage

ClusterX(
  data,
  dimReduction = NULL,
  outDim = 2,
  dc,
  gaussian = TRUE,
  alpha = 0.001,
  detectHalos = FALSE,
  SVMhalos = FALSE,
  parallel = FALSE,
  nCore = 4
)

Arguments

data	A data matrix for clustering.
dimReduction	Dimenionality reduction method.
outDim	Number of dimensions used for clustering.
dc	Distance cutoff value.
gaussian	If TRUE, apply gaussian to estimate the density.
alpha	Significance level for peak detection.
detectHalos	If TRUE, detect the halos.
SVMhalos	If TRUE, Run SVM on cores to assign halos.
parallel	If TRUE, run the algorithm in parallel.
nCore	Number of cores umployed for parallel compution.

Details

ClusterX works on low dimensional data analysis (Dimensionality less than 5). If input data is high dimensional, t-SNE is conducted to reduce the dimensionality.

Value

A list

Author(s)

Chen Hao

Examples

iris_unique <- unique(iris) # Remove duplicates
data <- as.matrix(iris_unique[,1:4])
ClusterXRes <- ClusterX(data)

JinmiaoChenLab/cytofkit2 documentation built on May 12, 2022, 8:09 a.m.