Network-Based Clustering.
clustNet: Network-based clustering with covariate adjustment
clustNet is an R package for network-based clustering of categorical data using a Bayesian network mixture model and optional covariate adjustment.
Installation
The package requires Rgraphviz and RBGL, which can be installed from Bioconductor as follows:
if (!requireNamespace("BiocManager", quietly = TRUE)) install.packages("BiocManager")
BiocManager::install(c("Rgraphviz", "RBGL"))
The latest stable version of clustNet is available on CRAN and can be installed with
install.packages("clustNet")
from within an R session. On a normal computer, this should take around 5-60 seconds, depending on how many of the required packages are already installed.
BiocManager::install("remotes")
Being hosted on GitHub, it is also possible to use the install_github tool from an R session to install the latest development version:
library("devtools")
install_github("cbg-ethz/clustNet")
clustNet requires R >= 3.5.
Example
library(clustNet)
# Simulate data
k_clust <- 3 # numer of clusters
ss <- c(400, 500, 600) # samples in each cluster
simulation_data <- sampleData(k_clust = k_clust, n_vars = 20, n_samples = ss)
sampled_data <- simulation_data$sampled_data
# Network-based clustering
cluster_results <- get_clusters(sampled_data, k_clust = k_clust)
# Load additional pacakges to visualize the networks
library(ggplot2)
library(ggraph)
library(igraph)
library(ggpubr)
# Visualize networks
plot_clusters(cluster_results)
# Load additional pacakges to create a 2d dimensionality reduction
library(car)
library(ks)
library(graphics)
library(stats)
# Plot a 2d dimensionality reduction
density_plot(cluster_results)
On a normal computer, the clustering should take around 2-4 minutes.