Dear all: i have the following problem I use gls, I have 10 repeated meassures. The 10 by 10 matrix has different variances in the diagonal (i know how to handle that) The correlation matrix is blocked: the first 3 repeated meassures have the same correlation, the next 4 have the same correlation and the last 3 the same one. and the correlation among the others is low. Is there any way to use a blocked structure in the correlation to fit this using gls? and if not, is there any way to fit using lme and random factors? Thansk, Lili -- Liliana Forzani Departamento de Matematica - Facultad de Ingenieria Quimica - Universidad Nacional del Litoral Instituto Matematica Aplicada Litoral - CONICET http://liliana.forzani.googlepages.com/english Tel: 54-342-4564339 Cel: 54-9-342-6134100 Trab: 54-342-4559175 (Int 2160) [[alternative HTML version deleted]]