Panel data- replacing missing individual values with the values of another individual, by time

[Greg Boudreaux]

Hi folks- I have a panel dataset that looks roughly like the example below. In my dataset, I have 5 years of observations, by day, for each individual. In this example, I've only included two months per year to keep it simple.
I'm trying to replace some individuals' missing values with the same time values as a specific other individual. In this example, indivual 3 is missing values of "measurement", and I want to replace these with the Individual 1's values of "measurement", matched by time. So, my current dataset looks like this:

* Example generated by -dataex-. To install: ssc install dataex
clear
input byte individual int year byte(month measurement)
1 2010 1  2
1 2010 2  9
1 2011 1  7
1 2011 2  5
1 2012 1  5
1 2012 2  4
2 2010 1 12
2 2010 2  2
2 2011 1 10
2 2011 2  7
2 2012 1  6
2 2012 2  3
3 2010 1  .
3 2010 2  5
3 2011 1  7
3 2011 2  .
3 2012 1  .
3 2012 2  9
end

And I need it to look like the example below. Any help would be greatly appreciated! Thanks.

* Example generated by -dataex-. To install: ssc install dataex
clear
input byte individual int year byte(month measurement)
1 2010 1  2
1 2010 2  9
1 2011 1  7
1 2011 2  5
1 2012 1  5
1 2012 2  4
2 2010 1 12
2 2010 2  2
2 2011 1 10
2 2011 2  7
2 2012 1  6
2 2012 2  3
3 2010 1  2
3 2010 2  5
3 2011 1  7
3 2011 2  5
3 2012 1  5
3 2012 2  9
end

[Rhys Williams]

Hi Greg,

Does something like the following work, it might require tweaking, I have shown how to replace 3 with 1 but I am not sure what the general rule you want to apply is:

bysort year month (individual): replace measurement = measurement[1] if individual==3

Best,
Rhys