Can anyone tell me where to find analogues of the Kolmogorov Smirnov test for distinguishing distributions in more than one dimension, or any other good way to test if two 2-dim distributions are different? I have two (small) data sets to compare, no theoretical distribution. I know about two papers in astrophysics on a generalisation of the KS test, I wondered if there are some in other fields. Thanks Alex Lewin %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%