I tried to run a GLM model using S-Plus and R. The same model and data were applied to the both packages. I got different results of P(>|Chi|) for Anova, between S-PLus and R. Other estimates values are the same. P(>|Chi|) for S-Plus: 0.99565 0.99682 0.04871 0.84597 1.00000 0.99999 1.00000 1.00000 1.00000 0.95834 P(>|Chi|) for R: 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00350 0.26400 0.46800 0.0000 I used S-Plus 6.1 for windows, Profesional Edition release 1, and R 1.8.1. Does anyone have any idea how this could happen? Thank you very much for your time. Ridwan [[alternative HTML version deleted]]
Prof Brian Ripley
2004-Mar-23 09:12 UTC
[R] Does Chi Square test for R differ from S-Plus?
Depends on the `GLM' model and what you mean by that (general linear model, generalized linear model?). And by `Anova' (capitalized), for that matter. S-PLUS (sic) does some inconsistent things with dispersion parameters in glm() fits, and it is likely that you are using the wrong test. But P values that near one should arouse your suspicions. Please read the R posting guide, and then tell us enough of what you did so we can help you. On Tue, 23 Mar 2004, Ridwan Sala wrote:> I tried to run a GLM model using S-Plus and R. The same model and data were > applied to the both packages. I got different results of P(>|Chi|) for Anova, > between S-PLus and R. Other estimates values are the same. > > P(>|Chi|) for S-Plus: > 0.99565 0.99682 0.04871 0.84597 1.00000 0.99999 1.00000 1.00000 1.00000 0.95834 > > P(>|Chi|) for R: > 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00350 0.26400 0.46800 0.0000 > > I used S-Plus 6.1 for windows, Profesional Edition release 1, and R 1.8.1. > > Does anyone have any idea how this could happen?-- Brian D. Ripley, ripley at 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