extracts a GMM from a posterior variational distribution. Only relevant components (i.e. associated to a significant population) are extracted.
boolean indicating wether to extract a label vector. If TRUE, model, a list object, should also contain a data attribute, used to build label vector.