Thread: k means algorithm question

Hi everyone! I'm new to the forum and I would like to receive some assistance on a problem that is troubling me. I've recently discovered the k-means algorithm for clustering, and I've been told that it is NP-hard for data dimensions higher than 2. My question is: how can I prove that the problem is not NP-hard when dimension is equal to 1 (points on a row)? The solution must have something to do with the fact that this kind of data can be sorted, but I can't find a mathematical proof. Any suggestion?