Clustering GPS data into "k" Groups

Question

I have a list of GPS(longitude and latitude pairs) data(~3000) and I would like to split them into "k" groups based based on their distance(geodesic and/or euclidean). What's the best way to do this?

Answer 1

Project the data into UTM.

Then choose an algorithm that you like. K-means if your data is very clean, for example.