As far as I'm aware, only the scoring logic produced by your model can be exported to EG. Anything to do with the model itself has to run within EM itself.   ... View more

Proc report builds things from top to bottom left to right. When you have this column statement: column var1 var2 maxvar1 minvar2; Then this compute compute var1; if var1 = maxvar1 then do; call define (_col_, "style", "style={background=peachpuff}"); end; endcomp; does not actually have a value of maxvar1 to use since maxvar1 is to the right of var1.   You might try changing the order of variables in the column statement so the maxvars come before var1 and var2.   ... View more

In the score data, you typically do not have an observed target, so you wouldn't be able to create an ROC curve for that. ... View more

Once you have the data in a SAS data set, the transformations are easy.  I could be wrong, but I believe that Enterprise Miner would use the same tools:   data want; set have; length short_cabin $ 1; short_cabin = cabin; lastname = scan(name, 1, ','); firstname_prefix = scan(name, 2, ','); run; ... View more