Cluster analysis (weird assignation)

Hi, I am trying to do some cluster analisys with kmeans () and fviz_cluster() and I dont understand why Switzerland is in the green cluster instead of in the purple one. Because clearly is closer to de purple? Any idea? Thanks

The short answer is that eigenvectors are being projected onto a Cartesian plane.

1 Like

There are more dimensions that are responsible for the cluster-assignement, here you only show Dim1 and Dim2, in here (I agree) Switzerland is close to cluster 4, however, maybe in dimension 3,4,5 it has a higher distance driving the separation into the used clusters.

You can check the other dimensions with the axes argument.