I am working on a linkage and I have two datasets with numerous missing values that are coded as all 9s in both datasets. I am trying to make sure that when the linkage codes for "Var_A" that it does not label matches for those that are all 9s. I want the other variables to match where applicable without matching those missing values. The example code is below. How might I edit that to accomplish this task? proc sql; create table Set_A as select a.Var_A, b.Var_A as Var_A, a.Var_B, b.Var_B as Var_B, a.Var_C, b.Var_C as Var_C from File1 a, File2 b where (complev(a.Var_B, b.Var_B, 'ILN') lt 1) and /*Perfect match on Var_B*/ ( ( b.Var_A =* a.Var_A and b.Var_B =* a.Var_B) and ((a.Var_C=b.Var_C)) ); quit;
... View more
