|From||"Moran, John (NWAHS)" <John.Moran@nwahs.sa.gov.au>|
|Subject||st: RE: Macro for goodness of fit|
|Date||Tue, 30 Aug 2005 09:45:42 +0930|
Perhaps "hl" is what you need
net from http://www.sealedenvelope.com/
and follow the links
Does anyone know of a do-file that can do a HL goodness of fit using only the predicted probability and actual probability? lfit will not work for me since I need to do the analysis in SAS (need multiple imputation) and I have more than more interaction term making it hard to import the predictors and use a simple covariate matrix.
Any help would be appreciated. Thanks.
Randall S. Burd, M.D., Ph.D.