Hello Axe, A do loop can be used to iterate through or clauses: [pre] %macro a; data Target; Set Source; if column = "&var1" %do i=1 %to &N; or column = "&&var&i" %end; then do; ... end; run; %mend a; %a [/pre] Sincerely, SPR ... View more

Hello Khill, The problem with your code is that all macro %IFs are resolved at compilation phase when AVG_Denom is not yet accessible. Reducing %IFs to datastep IFs solves the problem: [pre] %macro t; proc SQL noprint; select t1.AVG_Denom into :xAvgDenom from WORK.testdata2 as t1 ;quit; %put Average is &xAvgDenom; data _null_; if &xAvgDenom < 25 then put "&xAvgDenom is Less Than 25"; else put "&xAvgDenom is Greater Than 25"; run; %mend t; %t; [/pre] Sincerely, SPR ... View more

Hello Anandbilava, As AtrC noticed you have a truncation error in your character variables c3-c5 in the a and b datasets. If you add the line length c1 8 c2 c3 c4 c5 $3; after data a; and data b; then everything should work fine. At least it is true for my computer. Sincerely, SPR ... View more

Hello Axe, It seems to me that using an array of macro variables could help to solve the problem of "32k" but the program becomes more complicated. I mean using [pre] proc SQL; select count(Distinct ID) :into N from test; %let N=%TRIM(&N); select distinct ID into :ID1-:ID&N from TEST; quit; [/pre] with subsequent cycle on ID1 to ID&N. Sincerely, SPR ... View more

Hello IamJeannie, If you are not against coding and SQL in EG, you can use the following code: [Pre] data a; input Month $1 Is_Yes; datalines; a 1 a 0 a 1 b 1 b 0 ; run; /* Result */; proc SQL; create table r as select Month, 1-SUM(Is_Yes)/COUNT(*) as Is_Yes_by_Month format=Percent7.2, COUNT(*) as N from a group by Month ;quit; [/Pre] Sincerely, SPR ... View more

Sorry, part of the code has been cut by forum software. It does not like apostrophes. This is the last proc SQL: proc SQL; create table r as select ID, yr, wk, Count(*) as n, case when calculated n > 3 then 'n GT 3' else 'n LE 3' end as result from a1 group by ID, yr, wk ;quit; SPR ... View more

Hello Evolod, This is a possible solution to your problem. For the data you supplied all patients took less than 3 medications a week. data a; input ID 1-6 DRUG $ 8-14 DATE YYMMDD11.; format date date7.; datalines; 100001 ADVIL 2010-03-25 100001 TILENOL 2010-04-08 100001 TILENOL 2010-04-10 100001 TILENOL 2010-04-12 100001 TILENOL 2010-04-14 100001 TILENOL 2010-04-28 100001 TILENOL 2010-05-01 100002 ADVIL 2010-04-01 100002 ADVIL 2010-04-09 100002 ADVIL 2010-04-10 100002 ADVIL 2010-04-12 100002 ADVIL 2010-04-14 100002 TILENOL 2010-04-28 100002 TILENOL 2010-05-01 ; run; /* Add Year, Week */; data a1; set a; yr=YEAR(Date); wk=WEEK(Date); run; /* Check if patient is taking meds more that 3 times a week */; proc SQL; create table r as select ID, yr, wk, Count(*) as n, case when calculated n > 3 then ' >3' else '<=3' end as result from a1 group by ID, yr, wk ;quit; Sincerely, SPR ... View more

Hello Aravind, A functional style macro is explained in http://www2.sas.com/proceedings/sugi31/042-31.pdf Sincerely, SPR ... View more

Hello Sid S, If I correctly understood what you need, the following code produces the result: data a; input Name $ 1-4 Action $ 6-10 Quantity 11-13 Maturity $ 15-20; datalines; Joe Buy 100 Dec 12 Joe Buy 100 Dec 10 Joe Sell 50 Dec 10 Mary Buy 50 Dec 12 ; run; /* Add minus sign to quantity for activity=sell */; data a1; set a; if UPCASE(Action)="SELL" then quantity=-quantity; run; /* Summarizing Quantity by Name and Maturity */; proc SQL; create table r as select Name, SUM(quantity) as Quantity, Maturity from a1 group by Name, Maturity ;quit; Sincerely, SPR ... View more

Hello Enginemane44, It looks like your input dataset should contain an additional variable, say ID, to distinguish two sets of data. Anyway, to solve the problem for the first set of questions you can use the following code: data a; input QNUM Rating; datalines; 1 3.0 2 2.9 3 10.0 5 7.7 6 8.9 7 9.1 9 10.0 10 7.6 ; run; data t; do qnum=1 to 10; output; end; run; data r; merge a(in=a) t(in=t); if a and t then output; if not a and t then output; by qnum; run; Sincerely, ... View more

Sorry, I did not realize that you need the month from signdate. In this case the first data step should be replaced with the code: data r; set b; lm=(MONTH(logindate)+12*YEAR(logindate))- (MONTH(signdate )+12*YEAR(signdate )); run; SPR ... View more

Hello Kdang, Try the following code, where b is the data set containing id signupdate logindate activity: /* Add Login Month */; data r; set b; lm=MONTH(logindate); run; /* Count Start and Open Activities */; proc SQL; create table rr as select id, lm, activity, Count(*) as n from r group by id, lm, activity ;quit; /* Count Login Activities */; proc SQL; create table login as select id, lm, "login" as activity , sum(n) as n from rr group by id, lm ;quit; /* Combine results */; data rrr; set rr login; run; proc sort data=rrr; by id lm activity; run; Sincerely, SPR ... View more

Hello TMorville, SAS help says that it is necessary to use "V" or "W" descriptor to switch to Monday, say: Week("01Feb2010"d,"W"); Sincerely, SPR ... View more

I meant the library displayed on the top of library portion of the Server list. You can access the Server List by selecting View -> Server. Then double click on Servers, then of the server which you use, then on Libraries, and you will see this library list. ... View more

Hello WhiteSalmon, From my experience with EG, it is a good idea to have an underscore as a first character of a user library name. In this case this library is always displayed on the top of the library list. Sincerely, SPR ... View more