Dear Yumin,
This is least-elegant-but-work method.
1. If you get such data when you "insheet",
+----------------------------------------------+
| v1 v2 v3 v4 v5 |
|----------------------------------------------|
1. | country variable 1960 1961 2002 |
2. | Afganistan Aid 1 2 3 |
3. | Afganistan Population 4 5 6 |
4. | Afganistan Wage 7 8 9 |
5. | Albania Aid 10 11 12 |
|----------------------------------------------|
6. | Albania Population 13 14 15 |
7. | Albania Wage 16 17 18 |
8. | China Aid 19 20 21 |
9. | China Population 22 23 24 |
10. | China Wage 25 26 27 |
+----------------------------------------------+
foreach v of varlist v3 - v5 {
local a = `v'[1]
rename `v' v`a'
}
foreach v of varlist v1 - v2 {
local a = `v'[1]
rename `v' `a'
}
drop in 1
(1 observation deleted)
list
/*Result was omitted by me. I recommend to use list
to see what occurs by lines below.*/
2. Reshape yearly data vertically.
gen id = _n
gen vid = mod(id - 1,3) + 1
/* This variable "vid" is for the next step. */
reshape long v, i(id) j(year)
3. Reshpe with holizontal "variable id".
sort country year variable
replace id = _n
bysort country year: gen cy = _n
gen cyc = 1 if cy == 1
gen cyid = sum(cyc)
drop id variable cy cyc
reshape wide v, i(cyid) j(vid)
If you need it, you must rename v1-v3
with "aid", "population", and "wage".
> > I have been trying to save data from World
> > Development
> > Indicators (WDI, online) into a cross-section
> > time-series format. The original data look like the
> > following
> >
> > Country Variable Name 1960 1961 ?2002
> > Afghanistan Aid X X ?X
> > Afghanistan Population X X $B)9(J
> > ??????> ??????> Afghanistan Wage X X ?X
> > Albania Aid X X ?X
> > Albania Population X X ?X
> > ??????> ??????> Albania Wage X X ?X
> >
> > The ideal format I would like to have would be:
> >
> > Country Aid Population ??Wage
> > Afghanistan 1960 X X ?X
> > Afghanistan 1961 X X ?X
> > ??????> ??????> Afghanistan 2002 X X ?X
> > Albania 1960 X X ?X
> > Albania 1961 X X ?X
> > ??????> ??????> Albania 2002 X X ?X
> > ??????> China 1960 X X ?X
> > China 1961 X X ?X
> > ??????> ??????> China 2002 X X ?X
> >
> > The problem is that some countries have data on more
> > variables than do other countries. Any kindly
> > suggestion on how to execute the data manipulation
> > trick in stata or on how to save WDI online properly
> > would be tremendously appreciated. Many thanks!
> >
> > Best,
> > Yumin
> >
> >
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
