The goal of concordexR is to replace UMAP as a clustering diagnostic.
Installation
You can install the development version of concordexR from GitHub with:
# install.packages("devtools")
devtools::install_github("pachterlab/concordexR")Example
This is a basic example using concordex:
library(concordexR)
library(BiocNeighbors)
g <- findKNN(iris[, seq_len(4)], k = 10)
#> Warning in (function (to_check, X, clust_centers, clust_info, dtype, nn, :
#> detected tied distances to neighbors, see ?'BiocNeighbors-ties'
res <- calculateConcordex(g$index, labels = iris$Species, k = 10, return.map = TRUE)
plotConcordexSim(res)
heatConcordex(res)