Help changing missing value to the row mean

[rauzim]

I am trying to write code to handle missing items from survey data. The survey is the AM-PAC outpatient basic mobility (contains 18 questions), and per the survey instructions, to score surveys with missing data, you are supposed to find the average score for the survey's non-missing items, and then input that average score for the missing items.

Here is a snapshot of my data:

So you can see in obs 5 (pid SP_002), there are 3 missing items (highlighted in yellow); I've added a variable for the row mean (ampac_mean), and this rows mean is the value highlighted in yellow in the last column. 

How do I input that value (2.13) for the missing items in that row? 

I have additional rows with missing data so I will need to be able to input that row's mean value into the missing items. 

Thank you,

Michelle

[fja]

Hello!
As it appears to me that you do not want to calculate the values yourself, why not having an array of values and looping over it ... checking for . ?
Regards

[fja · Accepted solution]

data work.TestData;
	input Test09-Test11 Mean;
	datalines;
	1 2 3 2
	. 4 5 4.5
	6 . 8 7
	8 9 . .
	;
run;

data work.TestDataImputed;
	set work.TESTDATA;
	array Cols(*) Test09-Test11;
	
	if not missing(Mean) then do;
		/* Loop over cols */
		do i=1 to dim(Cols);
			if missing(Cols(i)) then Cols(i) = Mean;
		end;
	end;	
run;

or something similar ...