Hierarchical clustering with R

Question

Consider several points:

A = (1, 2.5), B = (5, 10), C = (23, 34), D = (45, 47), E = (4, 17), F = (18, 4)

How can I perform hierarchical clustering on them with R?
I've read this example Cluster Analysis but I'm not sure how to enter these values as points rather than just regular numbers.

When I do

x <- c(...) #x values
y <- c(...) #y values

I can plot them using

plot(x,y)

But how can I specify those values like in the example:

mydata <- scale(mydata)

Doing

mydata <- scale(x,y)

I get the following error

Error in scale.default(x, y) : 
  length of 'center' must equal the number of columns of 'x'

jlhoward · Accepted Answer · 2014-11-08T15:53:52.730

3

Something like this??

A = c(1, 2.5); B = c(5, 10); C = c(23, 34)
D = c(45, 47); E = c(4, 17); F = c(18, 4)
df <- data.frame(rbind(A,B,C,D,E,F))
colnames(df) <- c("x","y")
hc <- hclust(dist(df))
plot(hc)

This puts the points into a data frame with two columns, x and y, then calculates the distance matrix (pairwise distance between every point and every other point), and does the hierarchical cluster analysis on that.

We can then plot the data with coloring by cluster.

df$cluster <- cutree(hc,k=2)    # identify 2 clusters
plot(y~x,df,col=cluster)

enter image description here

edited Nov 08 '14 at 15:53

answered Nov 08 '14 at 15:27

jlhoward

58,004
7
97
140

Is it possible to specify the linkage criteria, like single or complete linkage? – Evgenij Reznik Nov 08 '14 at 23:45
1

Read the documentation on `hclust(...)`, with particular attention to the `method=...` argument (type `?hclust` at the R command line). – jlhoward Nov 08 '14 at 23:50

Hierarchical clustering with R

1 Answers1

Linked