I wish someone can help with this. I am using proc reg to model something. I used the exactly the same set of the variables in the model, but slightly changed the order of the variable. For example, let gender to be last variable in the model rather than the first one. The results i got is a bit different, althought, the differences are really small like 1.3081E-09.
Does anyone know why?
For example, my code:
proc reg data = hrp;
model tobsprob = Variable list / p;
reweight othjoin>=1 / weight=0;
output out=hrpmodel ;