inst/test_order_fa_univariate.R

In jchiquet/fusedanova: Univariate hierarchical clustering

rm(list=ls())
library(univarclust)

set.seed(111)
n1 <- 100; n2 <- 200; n3 <- 150; n <- n1 + n2 + n3
mu1 <- 2; mu2 <- 0; mu3 <- -2
sigma <- .1
trait <- setNames(
  c(rnorm(n1, mu1, sigma), rnorm(n2, mu2, sigma), rnorm(n3, mu3, sigma)),
  nm = paste0("ind",1:n)
  )
group <- factor(rep(paste0("grp",1:3), c(n1,n2,n3)))

# sampling ther data order
o <- sample(1:(n1 + n2 + n3))
group <- group[o]
trait <- trait[o]

fa <- univarclust::clusterpath_l1(trait, gamma = 0)

plot(fa, main = "fused-ANOVA tree")