Hello, I need to assign an analysis flag based partly on this "If a subject has both "High" and "Low" (in ANRIND) within a parameter, set to "Y" for the earliest occurrence of each value.". I am using PROC SQL to achieve this using the following code.   proc sql; create table want as select *, count(wrstdate) as count from have where wrstflg is not null group by subjid,paramcd; quit; This code is not overly important to understand but I'm trying to add this line of code   sum (distinct anrind in ('High','Low') as hlcnt) after count so I get a hlcnt of 2 so I can use it to keep these records when assigning a flag further downstream in the code. However, when I try this it partially works but it also keeps records when there is multiple values of high in different visits or same with multiple values of low which I don't want. I only want to keep it for subjects that have distinct value of high and low.   data have; input SUBJID $ PARAMCD $ VISIT :$30. ANRIND $ WRSTDATE $ WRSTCASE :3. WRSTFLG $; infile datalines dlm = '|'; datalines; 1001|AAA|Week 2 |Low|11JUL2022|2|Y 1001|AAA|Week 16|High|11JUL2022|3|Y 1002|BBB|Week 2 |Low|1AUG2022|2|Y 1002|BBB|Week 4 |Low|15AUG2022|2|Y 1003|CCC|Week 2 |High|1AUG2022|3|Y 1003|CCC|Week 4 |High|15AUG2022|3|Y ; run; data want; input SUBJID $ PARAMCD $ VISIT :$30. ANRIND $ WRSTDATE $ WRSTCASE :3. WRSTFLG $ HLCNT :3.; infile datalines dlm = '|'; datalines; 1001|AAA|Week 2 |Low|11JUL2022|2|Y|2 1001|AAA|Week 16|High|11JUL2022|3|Y|2 1002|BBB|Week 2 |Low|1AUG2022|2|Y|1 1002|BBB|Week 4 |Low|15AUG2022|2|Y|1 1003|CCC|Week 2 |High|1AUG2022|3|Y|1 1003|CCC|Week 4 |High|15AUG2022|3|Y|1 ; run;   ... View more

Hello I have this code that is doing as intended but I get this warning error in my log. How do I workaround that? I basically want maxrecdate and maxrec to merge to vischk1 by usubjid and adt for further processing but without it altering the current version of my data in totedps1_pre with the merge statement applied proc sql; create table maxrec1 as select distinct usubjid, adt, adt as maxrecdate, aval as maxrec, visit, from vischk1 where aval ne . and visit not in('Screening','Day 1','Week 1') and max(swrel,swnrel,t2cm,rtt1l28d)=0 order by usubjid,paramcd,adt,visit; run; data totedps1_pre; merge vischk1 maxrec1; by usubjid adt; run;   ... View more

Hello, I have a invalue format to populate the value of VISITNUM for some records where it's value is missing using AVISITN which is populated for all records as a reference. Both are numeric variables. So if AVISITN = 10 I would want the missing VISITNUM  to be 1 etc.   data have; input VISITNUM :8. AVISITN :8.; infile datalines dlm = '|'; datalines; 1|10| 2|20| 4|30| 5|40| 6|50| 7|60| 8|70| 1|10| 2|20| 4|30| 5|40| 6|50| |60| |70| |80| 1|10| 2|20| 4|30| 5|40| |50| |60| |70| |80| 1|10| 2|20| |30| |40| |50| |60| |70| |80| ; RUN; data want; input VISITNUM :8. AVISITN :8.; infile datalines dlm = '|'; datalines; 1|10| 2|20| 4|30| 5|40| 6|50| 7|60| 8|70| 1|10| 2|20| 4|30| 5|40| 6|50| 7|60| 8|70| 9|80| 1|10| 2|20| 4|30| 5|40| 6|50| 7|60| 8|70| 9|80| 1|10| 2|20| 4|30| 5|40| 6|50| 7|60| 8|70| 9|80| ; RUN;     proc format; invalue dummy_visnum 10 = 1 20 = 2 30 = 4 40 = 5 50 = 6 60 = 7 70 = 8 80 = 9 100= 10; quit; However when I run this code it works as intended but I get a warning in my log "Numeric values have been converted to character values at the places given by: (Line):(Column)."       visitnum = input(avisitn, dummy_visnum.); What is the workaround to eliminate this?    ... View more