If I have outliers in my variable. How can I program SAS to find those outliers and make them the average of the previous value and the subsequent value? For example,
if malaria cases are 1714 for Feb 2018 I want SAS to take the average between Jan 2018 (200 cases) and March 2018 (47 cases).
Thank you.
Here's a way to average the prior and post values. You still have to decide what makes a value an outlier and incude that logic in the program.
data want;
merge have have (obs=2 keep=variable rename=(variable=subsequent));
prior = lag(variable);
if /* your logic determines that current VARIABLE value is an outlier */ then do;
if _n_=1 then capped_value = subsequent;
else capped_value = mean(subsequent, prior);
end;
else capped_value = variable;
run;
Then use CAPPED_VALUE instead of VARIABLE.
Here's a way to average the prior and post values. You still have to decide what makes a value an outlier and incude that logic in the program.
data want;
merge have have (obs=2 keep=variable rename=(variable=subsequent));
prior = lag(variable);
if /* your logic determines that current VARIABLE value is an outlier */ then do;
if _n_=1 then capped_value = subsequent;
else capped_value = mean(subsequent, prior);
end;
else capped_value = variable;
run;
Then use CAPPED_VALUE instead of VARIABLE.
Thank you. Can you please explain what datasets I would put in the 2 'have' part and what 'keep 2 observations' means?
What can be place in the logic if I want the outlier to be +/- 3 SD from mean?
"Have" is the name of the data set that you have that contains all the data.
The obs=2 tells SAS to start reading that data set with the second observation. So one "have" starts at the first observation, and the other "have" reference starts at the second observation, thus getting the subsequent value for your variable.
Registration is open! SAS is returning to Vegas for an AI and analytics experience like no other! Whether you're an executive, manager, end user or SAS partner, SAS Innovate is designed for everyone on your team. Register for just $495 by 12/31/2023.
If you are interested in speaking, there is still time to submit a session idea. More details are posted on the website.
Learn the difference between classical and Bayesian statistical approaches and see a few PROC examples to perform Bayesian analysis in this video.
Find more tutorials on the SAS Users YouTube channel.