I am trying to fit the generalised linear model for the negative binomial, but the results which come out are attached below. When we fit this model using few covariates, the model converge. Does it mean that this family is fitted differently from other glm? or the number of zeros in my response variable has a limiting factor? Thanks Bruno fit <- glm.nb(pfde~SEX+...., data=data1) Warning messages: 1: Algorithm did not converge in: glm.fitter(x = X, y = Y, w = w, etastart = eta, offset = offset, 2: Algorithm did not converge in: glm.fitter(x = X, y = Y, w = w, etastart = eta, offset = offset, 3: Algorithm did not converge in: glm.fitter(x = X, y = Y, w = w, etastart = eta, offset = offset, 4: Algorithm did not converge in: glm.fitter(x = X, y = Y, w = w, etastart = eta, offset = offset, 5: Algorithm did not converge in: glm.fitter(x = X, y = Y, w = w, etastart = eta, offset = offset, 6: Algorithm did not converge in: glm.fitter(x = X, y = Y, w = w, etastart = eta, offset = offset, 7: Algorithm did not converge in: glm.fitter(x = X, y = Y, w = w, etastart = eta, offset = offset, 8: alternation limit reached in: glm.nb(pden ~ SEX + RES + TRAVEL + TRAVHI + --------------------------------- Want to chat instantly with your online friends? Get the FREE Yahoo!Messenger [[alternative HTML version deleted]]