Re: st: wide to long

[David Kantor <kantor.d@att.net>]

Hi Alice,

I'm not totally sure of where the message is coming from, but I would 
ask, what variables are present when you do the -reshape long- ?

If I am not confused, year (in your example) should not exist prior 
to the reshape.  It is created by -reshape long-. I don't see how you 
would have a year variable in the wide form (as your message 
suggests); what would it signify?

The variables you should have prior to -reshape wide- ought to be
 firmid
 totistam1985
 totistam1986
 totistam1987
  (etc. -- other totisam variables)

And that's all (at least, in your example, doing just one measure at a time).

By the way, any other variables not mentioned in the -reshape- 
command, but which exist in your data, must be (I believe) 
functionally dependent on firmid.  But in your case, there really 
ought not to be any other variables.

I hope this helps.
--David

At 10:56 AM 10/19/2006, you wrote:

> Hi David, thanks for the response. I transfered each individual variable
> file to State using Stat transfer. I renamed the variable using a prefix.
> However, when I try:
>
> reshape long totistam, i(firmid) j(year)
>
> I get the following message:
>
> year contains all missing values
> r(498);
>
> Could anybody please tell me what's going on? I also tried - mvencode - to
> replace missing values with 0 but that did not help either.
> Best,
> alice
>
> [...]
*
*   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/