I am trying to plot a k means cluster plot in R - r

Question

I am trying to plot a k means cluster for the following data set:
Elo Rank Elo Score
Man City 1208
Man United 1123
Tottenham 1121
Liverpool 1107
Chelsea 1064
Arsenal 1032
Crystal Palace 996
Burnley 992
Everton 988
Bournemouth 978
West Ham 976
Newcastle 970
Leicester 965
Brighton 955
Southampton 938
Watford 927
Huddersfield 926
West Brom 919
Stoke 914
Swansea 901
I am trying to run kmeans code but Im getting the error:
'Error in colMeans(x, na.rm = TRUE): 'x' must be numeric'
Im assuming this is because of the first column. however I want to label each point on the plot with the team name so I know which point is what.
an example of what I want is the first diagram in this link:
https://www.geeksforgeeks.org/clustering-in-r-programming/
how do i go about plotting this?

Answer 1

Maybe you want something like this. I created 4 clusters of your data using kmeans. You can use this code:
First your data:
# A tibble: 20 × 2
`Elo Rank` `Elo Score`
<chr> <dbl>
1 Man City 1208
2 Man United 1123
3 Tottenham 1121
4 Liverpool 1107
5 Chelsea 1064
6 Arsenal 1032
7 Crystal Palace 996
8 Burnley 992
9 Everton 988
10 Bournemouth 978
11 West Ham 976
12 Newcastle 970
13 Leicester 965
14 Brighton 955
15 Southampton 938
16 Watford 927
17 Huddersfield 926
18 West Brom 919
19 Stoke 914
20 Swansea 901
Next scale your data (really important for kmeans clustering):
df$`Elo Score` <- scale(df$`Elo Score`)
Next create 4 clusters:
library(factoextra)
# Compute k-means with k = 4
km.res <- kmeans(df$`Elo Score`[!is.na(df$`Elo Score`)], 4)
Results:
print(km.res)
K-means clustering with 4 clusters of sizes 9, 1, 4, 6
Cluster means:
[,1]
1 -0.1962215
2 2.4819362
3 1.2379850
4 -0.9446472
Clustering vector:
[1] 2 3 3 3 3 1 1 1 1 1 1 1 1 1 4 4 4 4 4 4
Within cluster sum of squares by cluster:
[1] 0.5729761 0.0000000 0.3216049 0.1143089
(between_SS / total_SS = 94.7 %)
Available components:
[1] "cluster" "centers" "totss" "withinss" "tot.withinss" "betweenss" "size" "iter"
[9] "ifault"
Finally add your clusters to data:
# add clusters to data
df$cluster <- km.res$cluster
Result:
# A tibble: 20 × 3
`Elo Rank` `Elo Score`[,1] cluster
<chr> <dbl> <int>
1 Man City 2.48 2
2 Man United 1.47 3
3 Tottenham 1.44 3
4 Liverpool 1.28 3
5 Chelsea 0.764 3
6 Arsenal 0.382 1
7 Crystal Palace -0.0477 1
8 Burnley -0.0955 1
9 Everton -0.143 1
10 Bournemouth -0.263 1
11 West Ham -0.286 1
12 Newcastle -0.358 1
13 Leicester -0.418 1
14 Brighton -0.537 1
15 Southampton -0.740 4
16 Watford -0.871 4
17 Huddersfield -0.883 4
18 West Brom -0.967 4
19 Stoke -1.03 4
20 Swansea -1.18 4