I would probably 1. add a new observation with values like 1,2,3,4,1,2,3,4,1,2,3,4,.... 2. use Proc Transpose When your dataset is large, this can be super efficiant by avoiding loops.
... View more
I prefer to use ods rtf. So you can get an editable output, which is very similar to word file. Try ods rtf style=statistical bodytitle; /* your procedure*/ ods rtf close; Bodytitle is an option to make nicer output. You can also use "style=journal" instead of "statistical" to get different style of output. Hope this can help you.
... View more
I think you can try to use SQL for this problem. Suppose dataset "lookup" is the one with 450 observations, and dataset "healthcare" is the large healthcare claims database you have. lookup has a variable called var1 healthcare has 4 variables called dx1-dx4 Then you can create a dataset "selected" outputing all selected observations from healthcare. You might need to replace with your own dataset name or variable names. Proc sql; create table selected as select * from healthcare where dx1 = any (select var1 from lookup) or dx2 = any (select var1 from lookup) or dx3 = any (select var1 from lookup) or dx4 = any (select var1 from lookup) ; quit;
... View more
The way you put referece level looks correct to me. I tested your code. You can find a table "Class Level Information" from the output (see below). The "0-0-0" in "Design Variables" represents the reference group, which is what you specified. Be careful about your formats. Better to remove those blanks from format. Sometime, this can cause unexpected problem and can be hard to debug. Proc Format; Value timef 1='6 months to less than 12 months ' 2='12 months to less than 24 months ' 3='24 months to less than 36 months ' 4='36 months and above' ; Value $oldf '0'='Young' '1'='Old' ; run;
... View more
Try this: Proc sort data=dat1 nodupkey out=dat1_updated; by Date Ord_no item update_date; run; Proc sort data=dat2 nodupkey out=dat2_updated; by Date Ord_no item update_date; run; data finaldat; merge dat2_updated(in=a) dat1_updated(in=b); by Date Ord_no item update_date; if a and not b; run; Proc sql can also be used for this scenario.
... View more
The same thing I was thinking. Just adding additional "distinct" and "order by" to clean up the output proc sql; select distinct V1 from have where V1 not in (select distinct V2 from have) and V1 not in (select distinct V3 from have) order by V1 ; quit;
... View more
* is a special character and you need to mask it when assigning to a macro variable. Try the macro function %STR to mask *. Not sure how your "%Let Interaction = A*B;" works, but try to use "%let Interaction=%str(A B A*B);". The same solution for your macro problem. See more details from thel link below https://v8doc.sas.com/sashtml/macro/z3514str.htm Hope it can help.
... View more
