MSclust.NB: Mean-shift clustering of circular data: non-blurring mean...

In ClairBee/AS.circular: ArchStats: circular functions

Description

Iterative function to cluster angular data about an unspecified number of modes.

Usage

MSclust.NB(data, p, conv.lim = 1e-05, max.runs = 20)

Arguments

data	Set of angular data to be clustered.
p	Stabilising parameter p, estimated using MSclust.p.est.
conv.lim	Required degree of convergence: when the degree of correlation between successive iterations is closer to 1 than this, the data is deemed to have converged. Default is 0.00001.

Value

Matrix containing all iterations of the data set.

Examples

ex1 <- c(rvonmises(120 * 0.3, mu = circular(pi/2), kappa = 10),
         rvonmises(120 * 0.7, mu = circular(pi), kappa = 3))
p <- MSclust.p.est(ex1, dp = 4)
nb <- MSclust.NB(ex1, p = 7, conv.lim = 0.000001)

ClairBee/AS.circular documentation built on Jan. 24, 2020, 3:57 p.m.