Simple SAS Questions

[NormalDan]

Originally Posted by LICENSED TO ILL

Im having trouble finding a way to fill in the blanks by using the previous line. Help.

MemberID	DATE	UNIT_FACTOR	MEDICAID_FLAG
12345	201801	 1.0000 	Y
12345	201802	 1.0000 	Y
12345	201803	 1.0000 	Y
12345	201804		
12345	201805		
12345	201806	 1.0000 	Y
98765	201711	 1.0250 	N
98765	201712	 1.0250 	N
98765	201713		
98765	201714	 1.0250 	N
98765	201715	 1.0250 	N
98765	201716	 1.0250 	N

Not sure if there's a fancier way, but what I'd do is create two tables, one with the value you're going to match with and the other would contain the blanks.

To get your match values you'd want to create a table of those records which don't have blanks, then sort by descending date and then dedupe on member ID. Then your other table is just a dataset of those who had blanks. Then match across your values from one table to the next to fill in your second table's blanks (being careful to drop extra variables in the first table so as not to overwrite them in the second).

Then recreate your original table by setting together those who originally did not have blanks and those who had blanks but now don't.

[Childish Gambino]

This is somewhat clunky but should work

data your_data2;
 set your_data;
 format unit_fill 8.4;
 format mcd_fill $1.;
 if unit_factor = . then unit_factor = unit_fill;
 if medicaid_flag = . then medicaid_flag = mcd_fill; 
 mcd_fill = medicaid_flag;
 unit_fill = unit_factor;
 retain mcd_fill unit_fill;
 drop mcd_fill unit_fill;
run;

[LICENSED TO ILL]

Originally Posted by Childish Gambino

This is somewhat clunky but should work

I wouldn't say clunky, just easy button. Thanks

This is what I tried but forgot the RETAIN

I hope this works