If we assume you have legal variable names, the following should work:
. l
+-------------------------------+
| code v890216 v691693 |
|-------------------------------|
1. | 5/15/1995 0 . |
2. | 5/16/1995 0 .1 |
3. | 5/17/1995 0 .2 |
4. | 5/22/1995 . .2 |
5. | 5/23/1995 0 .2 |
+-------------------------------+
. reshape long v , i(code) j(firm)
(note: j = 691693 890216)
Data wide -> long
-----------------------------------------------------------------------------
Number of obs. 5 -> 10
Number of variables 3 -> 3
j variable (2 values) -> firm
xij variables:
v691693 v890216 -> v
-----------------------------------------------------------------------------
. sort firm code
. l
+-------------------------+
| code firm v |
|-------------------------|
1. | 5/15/1995 691693 . |
2. | 5/16/1995 691693 .1 |
3. | 5/17/1995 691693 .2 |
4. | 5/22/1995 691693 .2 |
5. | 5/23/1995 691693 .2 |
|-------------------------|
6. | 5/15/1995 890216 0 |
7. | 5/16/1995 890216 0 |
8. | 5/17/1995 890216 0 |
9. | 5/22/1995 890216 . |
10. | 5/23/1995 890216 0 |
+-------------------------+
Scott
----- Original Message -----
From: Susan Flaherty <[email protected]>
Date: Tuesday, February 15, 2005 12:31 pm
Subject: st: Reshape Data
>
> Dear All,
> I have an unusual dataset that I am trying to reshape, and the ususal
> commands don't seem to apply.
> An example of the data is as follows:
>
> Code 890216(VO) 691693(VO)
> 5/15/1995 0 .
> 5/16/1995 0 0.1
> 5/17/1995 0 0.2
> 5/22/1995 . 0.2
> 5/23/1995 0 0.2
>
> I would like to obtain a standard panel (long format) so that I
> can run
> some regressions:
>
> FIRM DATE TURNOVER
> 890216 5/15/1995 0
> 5/16/1995 0
> 5/17/1995 0
> 5/22/1995 .
> 5/23/1995 0
> 691693 5/15/1995 .
> 5/16/1995 0.1
> 5/17/1995 0.2
> 5/22/1995 0.2
> 5/23/1995 0.2
>
>
> The only unique identifier is the firm number (890216(VO)), and the
> stem of the common text is the suffix rather than the prefix (e.g.
> inc_90 inc_91 inc_92).
>
> I am currently trying something like:
> . reshape datafile 890216, i(code) j(turnover) string
>
> But since the prefix is a unique number, I can't replace the unique
> number thousands of times.
> I would greatly appreciate any suggestions on how to reformat the
> data.
>
> Thank you and regards,
> Susan Flaherty
>
*
* 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/
