Please, I need help to change the dataset below (in the wide format) to
long format. Note that the reshape command is not enough to do this
transformation.
panel time schooling wage
1. 1 1996 15 3,800
2. 1 1996 16 3,500
3. 1 1996 14 3,850
4. 1 1997 15 3,999
5. 1 1997 16 4,000
6. 2 1997 15 4,258
7. 2 1998 16 4,854
8. 2 1998 14 4,555
9. 2 1998 16 4,159
10. 2 1998 16 4,369
11. 2 1998 15 4,758
12 3 1999 . .
13 3 1999 . .
14 3 1999 . .
15 4 2000 17 4,999
16 4 2000 17 4,870
17 4 2000 17 4,899
18 4 2000 16 5,000
19 4 2000 17 5,409
20 5 2001 17 5,398
21 5 2001 17 5,478
22 5 2001 16 5,369
***************************************************************
***** I need the data in this format (with the year controlling the time
series operator) ********************
panel time schooling wage
1. 1 1996 15 3,800
2. 1 1997 16 3,500
3. 1 1998 14 3,850
4. 1 1999 15 3,999
5. 1 2000 16 4,000
6. 1 2001 15 4,258
7. 2 1996 16 4,854
Frank replied that the posting did not indicate what Paulo had in mind.
That is surely the case; for one thing, both of these are in the long
format, so it is not a wide-to-long issue. Secondly, it is very strange
that the 6th observation (of 4258) should end up in panel 1, with its year
switched from 1997 to 2001, and the year of the seventh obs. should be
transformed from 1998 to 1996, etc. unless the issue is that the original
data are mislabeled--in which case reshape cannot help, nor can any Stata
command, if the data are incorrectly labelled. Paulo should provide a
description of what is to be accomplished here, with a small example
illustrating the before-after that follows that logic. Rewriting the data
to reflect the incorrect coding of observations is not a logical or Stata
programming issue.
