I would like to know how can I pot a k-media cluster solution, based on 7 variables. All codes I found on internet are based on two variables.

Thank you

# Plotting k-media cluster solution

Welcome to the community!

I suppose you meant k-median or k-medoid clustering, since as far as I know, there's nothing called k-media clustering.

If I understand your question correctly, you have some 7-dimensional observations, and you've performed clustering. Now, you want to plot the points and make some distinction (adding attributes in form of colour, size, shape, etc.) between points of different clusters. Is that right?

I don't think it's possible. The main obstacle will be plotting in \mathbb{R}^7. How do you plan to do that? Usually, plotting clustering in 2 or 3 dimensions are done by just plotting the points first with respect to the axes corresponding to the variables, and then change the attributes of the points according to the cluster they belong to. I'm afraid it won't be possible directly.

That said, one option will be to plot the points first in 3D plane with respect to 3 variables of your choice, and then add attributes according to the 4 other variables. Finally, add another attribute corresponding to the cluster. In theory, it'll work, but I'm very doubtful whether it 'll help in visualisation at all.

I hope I've explained my point of view, though rather pessimistic. Sorry!

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.