sdbscan: Simple density based 2d clustering.

In OuNao/sumapc: Sequential UMAP Clustering

Simple density based 2d clustering.

Description

Simple density based 2d clustering.

Usage

sdbscan(
  x,
  minpts = 100,
  maxlvl = 100,
  alfa = 0.05,
  bw = 0.25,
  nbins = 1,
  theta = 5,
  mvpratio = 0.5,
  search = c("first", "wider", "median"),
  ret_model = FALSE,
  plotcuts = FALSE
)

Arguments

x	a data matrix.
minpts	min cluster size.
maxlvl	max sequential level of space partitioning.
alfa	numeric. Exclude alfa portion of each extremity before search for a cut point in density curve.
bw	bin width.
nbins	min number of bins to search.
theta	integer. Angle of rotation in each step.
mvpratio	max valley/peaks value to allow cut.
search	Cut point search method. "first" returns the first cut valid cut point found, "wider" the point with max bins with no points and "median" the point closer to the center of the distribution.
ret_model	logical.
plotcuts	logical; plot data with histograms and cut point.

Value

A vector with cluster numbers. Length = nrow(x)

OuNao/sumapc documentation built on May 5, 2024, 12:35 p.m.