Hi, I am currently running into problem with proc means and big datasets. I am using proc means to calculate sums by group like this:  proc means data=x7 noprint; class id; var amount1-amount&end.; where id ne "" or id ne "."; output out=out1 sum=; run; My dataset x7 is approximately 70 million rows and the output should be around 3 million rows (the number of different id's). The procedure works correctly with a smaller dataset (around half million rows). However, the output with the bigger dataset is only 2 rows (sum row and some random id). Is there a more efficient way to do this? Or what is the problem here? I can not use proc sql as I want to calculate sums for a number of variables with the number not being fixed.  ... View more

Hi, I am looking to be able to set a custom interval and then create month dummies for observations with a macro. I am running into problems with the comparison of dates. In addition to what is below, I have also tried using quotes around the macro date variables and using the %sysevalf-function with %if-%then-%else -structure (and both at the same time). In the %sysevalf-case the date comparison is just always true or false. Below is my code:     %macro date_loop(start,end); data WORK.x7; set work.x99; %let start=%sysfunc(inputn(&start,anydtdte9.)); %let end=%sysfunc(inputn(&end,anydtdte9.)); %let dif=%sysfunc(intck(month,&start,&end)); %do i=0 %to &dif; %let date=%sysfunc(intnx(month,&start,&i,b),date9.); %let date2=%sysfunc(intnx(month,&start,&i,e),date9.); if mydate1 <= &date2. AND mydate2 >= &date. then %do; if employed = 'true' then p&i. = 1; if dummy = 'true' then p&i. = 1; %end; else %do; p&i. = 0; %end; %end; run; %mend date_loop; %date_loop(01jan2019,01jan2020); which results in following errors   ERROR 22-322: Syntax error, expecting one of the following: !, !!, &, *, **, +, -, /, <, <=, <>, =, >, ><, >=, AND, EQ, GE, GT, IN, LE, LT, MAX, MIN, NE, NG, NL, NOTIN, OR, ^=, |, ||, ~=. ERROR 388-185: Expecting an arithmetic operator. ERROR 76-322: Syntax error, statement will be ignored. I would be grateful for any help! ... View more