Data imputation in SAS

Reply
Regular Contributor
Posts: 196

Data imputation in SAS

How to do this in SAS:

I want to replace a missing value with the mean of lagged value and its next value?

Trusted Advisor
Posts: 1,147

Re: Data imputation in SAS

Posted in reply to munitech4u

i believe it could be achieved by using the lag function.

may be sample data and the expected output could help to get better solution.

Thanks,

Jag

Thanks,
Jag
Trusted Advisor
Posts: 3,215

Re: Data imputation in SAS

Posted in reply to munitech4u

There have been a lot of the same questions. https://communities.sas.com/thread/59681

Or google: "impute missings site:communities.sas.com"

---->-- ja karman --<-----
Esteemed Advisor
Posts: 5,521

Re: Data imputation in SAS

Posted in reply to munitech4u

If you have access to SAS/ETS, use the Expand procedure :

data myData;

input time value;

datalines;

0 -100

1 10

2 25

3 .

4 35

5 .

6 .

7 55

8 -1

;

proc expand data=myData out=filledData method=none;

id time;

convert value=filledValue / transformout=(missonly cmovave 3);

run;

proc print data=filledData noobs; run;

PG

PG
Ask a Question
Discussion stats
  • 3 replies
  • 208 views
  • 1 like
  • 4 in conversation