At each time point, I have a multivariate distribution $x=[x_1,x_2,x_3]$. I don't have that many time points around 3-15. I would like to know whether the change of the distribution over the time is significant. I checked out the Mann-Kendall trend test but it only works for a single value at each time point.

Do you guys know some good test for this? For a univariate distribution is fine with me as well.

  • $\begingroup$ You could look at changepoint analysis. $\endgroup$ – machineguyei May 27 '15 at 1:44
  • $\begingroup$ Thank you very much. I should be more precise since I am interested in more in monotonic trends. $\endgroup$ – Benjamin Nguyen May 27 '15 at 1:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.