A reparameterization to improve numerical optimization in multivariate REML (co)variance component estimation.docVIP

in multivariate (co)variance component estimation E Groeneveld Institute of Animal Husbandry and Animal Behaviour, Federal Agricultural Research Center, H?ltystr 10, 31535 Neustadt, Germany (Received 28 February 1994; accepted 15 June 1994) Summary - Multivariate restricted maximum likelihood (REML) (co)variance component estimation using numerical optimization on the basis of Downhill-Simplex (DS) or quasi- Newton (QN) procedures suffers from the problem of undefined ’covariance matrices’ as produced by the optimizers. So far, this problem has been dealt with by assigning ’bad’ function