[SOLVED] Copy value of partnered (infant) variable to original (mother)

Copy value of partnered (infant) variable to original (mother)

I have Dyadic (mother/infant) repeated measures data in long form.

I have three ID variables: Individual ID, Dyad ID, and "status".

ID	DYAD	Status	date	infant weight
001	01	0	01/01
101	01	1	01/01	10
001	01	0	02/02
101	01	1	02/02	20
002	02	0	01/01
102	02	1	01/01	11
002	02	0	02/02
102	02	1	02/02	21

I want to add infant weight to the mother's rows based on key variables: date and DYAD ID. So final results should look like:

ID	DYAD	Status	date	infant weight
001	01	0	01/01	10
101	01	1	01/01	10

Normally, I do it entirely through the GUI by creating a new mini-dataset by 1) selecting only infants (status ==1) and 2) with only key variables and variables of interest, delete infant weight from original and merge data sets: add variables based on key values.

This works fine, but I know there must be a way to do this with syntax.

Solution

You can do this by aggregating:

aggregate /outfile=* mode=addvariables overwritevars=yes
          /break=DYAD date /infant_weight=max(infant_weight).

Since in any pair of rows with the combined DYAD and Date there will be one row with a value in infant_weight and one row empty, the aggregate will fill the blank row with the maximum value of the pair - which is just the weight from the infant's row.