Hi There, I would like to estimate the parameters of a multi-variate t-distribution from a series of bivariate data. Can anyone suggest how this might be done in R? I typically use nlm() for fitting univariate distributions. Thanks, Ian -- Ian Jonsen, Postdoctoral Fellow Biology Dept., Dalhousie University, Halifax, NS CANADA B2H 4J1 office: 804 LSC voice: +1 902 494 3910 fax: +1 902 494 3736 email: jonsen at mathstat.dal.ca