After using kmeans(): how to determine which point belongs to which group?

Question

I am running a kmeans clustering to identify labeled data. I ran pca and then kmeans and got the following plot using ggbiplot:

Now, how can I determine which point belongs to which group in table format. That is, in my original data, I would like to label each point with its group.

Please include reproducible (sample) data and code to generate the plot. — Maurits Evers, Apr 17 '19 at 04:43
That does not look like good clusters to me... Are you trying to show that it doesn't work on this data? Why are you using 35 clusters when the projection shows there is only 1? — Has QUIT--Anony-Mousse, Apr 17 '19 at 18:22

score 1 · Answer 1 · answered Apr 17 '19 at 05:02

Assuming that the name of your dataframe is df and you want k clusters. After you run the k means function...

# K-Means CA
fit <- kmeans(df, k) # where k is the number of clusters

... you must include the groups that where produced from the fit into your dataframe

# add clusters to the dataframe
df$clusters <- fit$cluster
df
             a          b clusters
1  -0.96193342 -0.7447816        1
2  -0.29252572 -1.1312186        1
3   0.25878822 -0.7163585        1
4  -1.15213189  0.2526524        1
5   0.19578283  0.1520457        1
6   0.03012394 -0.3076564        1
7   0.08541773 -0.9530173        1
8   1.11661021 -0.6482428        2
9  -1.21885742  1.2243136        1
10  1.26736872  0.1998116        2

Data used in the example

set.seed(3)
n <- 10
k <- 2
df <- data.frame(a= rnorm(n), b= rnorm(n))

You can also have a look here.

After using kmeans(): how to determine which point belongs to which group?

1 Answers1