RV = RV.mle(X) computes the maximum likelihood estimate based
  the data X. X should be nObservations x nDimensions. X must
  be a count matrix, consisting of only zeros and ones. A one
  in the ith column indicates that the sample in the jth row is
  of class i.