Hi,
I have a given variable which I want to use to dynamically create two new variables. For context, this is related to calculating life tables from mortality data. Here's the starting point:
data test1; input a; datalines; 0.1 0.2 0.1 0.2 ; run;
The dataset I want to create is one where variable b starts at an arbitrary number (here 100), c is defined as a*b, and all the following values for b are lag(b)-lag(c), as follows:
data want; input a b c; 0.1 100 10 0.2 90 18 0.1 72 7.2 0.2 64.8 12.96 ; run;
I tried using retain and lag, but something is off as the assignment of c uses a from the same row and b from the previous row, whereas I would like it to use b from the same row as well. So here's my effort that doesn't work as intended:
data test2; set test; retain b 100; c = a*b; lag_c = lag(c); if lag_c ne . then b = b - lag_c; run;
The solution and my mistake is probably pretty basic, but having mostly an R background these functionalities are still somewhat unintuitive, so any help is welcome!
Hi @Alppen and welcome to the SAS Support Communities!
You can compute b after an OUTPUT statement and thus omit the LAG function:
data want;
set test1;
retain b 100;
c=a*b;
output;
b=b-c;
run;
Hi @Alppen and welcome to the SAS Support Communities!
You can compute b after an OUTPUT statement and thus omit the LAG function:
data want;
set test1;
retain b 100;
c=a*b;
output;
b=b-c;
run;
You don't need lag at all, retaining b and an output-statement is all you need:
data test2;
set test1;
retain b 100;
c = a * b;
output;
b = b - c;
run;
Edit: i really should have Firefox reload a page before starting to write an answer.
Build your skills. Make connections. Enjoy creative freedom. Maybe change the world. Registration is now open through August 30th. Visit the SAS Hackathon homepage.
Register today!Learn how use the CAT functions in SAS to join values from multiple variables into a single value.
Find more tutorials on the SAS Users YouTube channel.
Select SAS Training centers are offering in-person courses. View upcoming courses for: