compumap {RaceID}

Computation of a two dimensional umap representation

Description

This functions performs the computation of a two-dimensional umap representation based on the distance matrix in slot distances using the umap package.

Usage

compumap(
  object,
  dimRed = FALSE,
  n_neighbors = 15,
  metric = "euclidean",
  n_epochs = 200,
  min_dist = 0.1,
  local_connectivity = 1,
  spread = 1
)

Arguments

object

SCseq class object.

dimRed

logical. If TRUE then the umap is computed from the feature matrix in slot dimRed$x (if not equal to NULL). Default is FALSE and the umap is computed from the distance matrix stored in slot distances. If slot distances equals NULL dimRed is automatially set to TRUE.

n_neighbors

Umap parameter. See help(umap.defaults) after loading package umap. Default is 15.

metric

Umap parameter. See help(umap.defaults) after loading package umap. Default is "euclidean".

n_epochs

Umap parameter. See help(umap.defaults) after loading package umap. Default is 200.

min_dist

Umap parameter. See help(umap.defaults) after loading package umap. Default is 0.1.

local_connectivity

Umap parameter. See help(umap.defaults) after loading package umap. Default is 1.

spread

Umap parameter. See help(umap.defaults) after loading package umap. Default is 1.

Value

SCseq object with umap coordinates stored in slot umap.

Examples

sc <- SCseq(intestinalDataSmall)
sc <- filterdata(sc)
sc <- compdist(sc)
sc <- clustexp(sc)
sc <- findoutliers(sc)
sc <- compumap(sc)