Because "classIntervals" objects of style "hclust" or "bclust" contain hierarchical classification trees in their "par" attribute, different numbers of classes can be chosen without repeating the initial classification. This function accesses the "par" attribute and modifies the "brks" member of the returned "classIntervals" object.

getBclustClassIntervals(clI, k)
getHclustClassIntervals(clI, k)

Arguments

clI

a "classIntervals" object

k

number of classes required

Value

a "classIntervals" object with a "modified" attribute set

Author

Roger Bivand <Roger.Bivand@nhh.no>

See also

Examples

if (!require("spData", quietly=TRUE)) {
  message("spData package needed for examples")
  run <- FALSE
} else {
  run <- TRUE
}
if (run) {
data(jenks71, package="spData")
pal1 <- c("wheat1", "red3")
opar <- par(mfrow=c(2,2))
hCI5 <- classIntervals(jenks71$jenks71, n=5, style="hclust", method="complete")
plot(attr(hCI5, "par"))
plot(hCI5, pal=pal1, main="hclust k=5")
plot(getHclustClassIntervals(hCI5, k=7), pal=pal1, main="hclust k=7")
plot(getHclustClassIntervals(hCI5, k=9), pal=pal1, main="hclust k=9")
par(opar)
}

if (run) {
set.seed(1)
bCI5 <- classIntervals(jenks71$jenks71, n=5, style="bclust")
plot(attr(bCI5, "par"))
}
#> Committee Member: 1(1) 2(1) 3(1) 4(1) 5(1) 6(1) 7(1) 8(1) 9(1) 10(1)
#> Computing Hierarchical Clustering

if (run) {
opar <- par(mfrow=c(2,2))
plot(getBclustClassIntervals(bCI5, k=3), pal=pal1, main="bclust k=3")
plot(bCI5, pal=pal1, main="bclust k=5")
plot(getBclustClassIntervals(bCI5, k=7), pal=pal1, main="bclust k=7")
plot(getBclustClassIntervals(bCI5, k=9), pal=pal1, main="bclust k=9")
par(opar)
}