Please, tell what is the best way to create an R glm object with parameters etc estimated in SAS? I have a large dataset and bigglm fails to converge, so estimation is done is SAS. However, there are a lot of predictions that are much more easily done in R and I would like to use glm.predict and the flexibility of R. bigglm ends with an error, so no object is created. Thanks everybody. Stephen [[alternative HTML version deleted]]