Missing values are filled with the next observation

[HannaZhang]

Hi guys,

I want to fill Missing values with the next observation or previous observation under certain conditions.

The conditions as follows

The missing data at the first half  is replaced by the previous adjacent data, and the missing data at the second half is replaced by the next adjacent data. If there is residual data in the middle, the next adjacent data is used instead.

Thanks a lot.

Data have;
 input id var1;
 datalines;
 1 54
 1 60
 1 62
 1 .
 1 .
 1 .
 2 100
 2 100
 3 .
 3 .
 3 .
 3 .
 3 72
 ;
Run;

Data want;
 input id var1;
 datalines;
 1 54
 1 60
 1 62
 1 62
 1 100
 1 100
 2 100
 2 100
 3 100
 3 100
 3 72
 3 72
 3 72

 ;
Run;

[PeterClemmensen]

How bog is your data set?

[HannaZhang]

About 117mb