How to replace values of a variabel based on certain conditions?

[Titir Bhattacharya]

Hi,

pfa the example data:

* Example generated by -dataex-. For more info, type help dataex
clear
input float(id round pcd open_during) str1 firm
1 1 2 0 "a"
1 2 2 0 "a"
1 3 3 0 "a"
1 4 4 1 "a"
1 5 4 1 "a"
2 1 3 0 "b"
2 2 3 0 "b"
2 4 4 1 "b"
2 5 5 1 "b"
3 1 1 0 "c"
3 2 2 0 "c"
3 3 2 0 "c"
3 5 2 0 "c"
4 1 1 0 "d"
4 2 1 0 "d"
4 5 1 0 "d"
end

the data here shows postcodes(pcd) for each individual (id) in each round (round) and whether it opened during our period of study (open_during=1), and the firm they are employed in (firm). For each individual, I want to replace a postcode with the one they had before the postcode was opened, if their firm of employment has remained unchanged.

So my final data should look like following:

* Example generated by -dataex-. For more info, type help dataex
clear
input float(id round pcd open_during) str1 firm
1 1 2 0 "a"
1 2 2 0 "a"
1 3 3 0 "a"
1 4 3 1 "a"
1 5 3 1 "a"
2 1 3 0 "b"
2 2 3 0 "b"
2 4 3 1 "b"
2 5 3 1 "b"
3 1 1 0 "c"
3 2 2 0 "c"
3 3 2 0 "c"
3 5 2 0 "c"
4 1 1 0 "d"
4 2 1 0 "d"
4 5 1 0 "d"
end

I'm struggling with coding this and would appreciate any help.
Thanks!

[Andrew Musau]

This will do what you want with your example data:

bys id (round): replace pcd= pcd[_n-1] if firm==firm[_n-1] & open

But I wonder what happens if someone changes firm after a new postcode is open?

* Example generated by -dataex-. For more info, type help dataex
clear
input float(id round pcd open_during) str1 firm
1 1 2 0 "a"
1 2 2 0 "a"
1 3 3 0 "a"
1 4 3 1 "a"
1 5 3 1 "b"
end

Do you keep the new postcode then for that last observation (highlighted)?