Full_Name: David Clayton Version: 2.0.1 OS: Linux Submission from: (NULL) (131.111.126.242) Setting contrasts for a factor to be used in a model by contrasts(fact, how.many) <- mat where mat is a matrix does not coerce mat into storage mode "double". The resultant model.matrix is garbage.
contrasts (PR#7695)
2 messages · David Clayton, Brian Ripley
Do you have an example we can use to test a fix? I can only guess at what you supplied for `mat'. R doesn't really have `storage mode': and "single" should be fine. I guess (but only guess) you had an integer matrix.
On Fri, 18 Feb 2005 david.clayton@cimr.cam.ac.uk wrote:
Full_Name: David Clayton Version: 2.0.1 OS: Linux Submission from: (NULL) (131.111.126.242) Setting contrasts for a factor to be used in a model by contrasts(fact, how.many) <- mat where mat is a matrix does not coerce mat into storage mode "double". The resultant model.matrix is garbage.
Brian D. Ripley, ripley@stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UK Fax: +44 1865 272595