Sounds like
bysort hh firm time : replace value = sum(value)
by hh firm time : replace value = value[_N]
But this is safer:
bysort hh firm time : gen sumvalue = sum(value)
by hh firm time : replace sumvalue = sumvalue[_N]
or
egen sumvalue = total(value), by(hh firm time)
Nick
[email protected]
Alexander Staus
> I want to combine different observations in a panel dataset.
> e.g. for hh 72 (=id) I want to sum
> the variable "value" if the firm is the same on one date, but
> other variables just stay as they
> are.
>
> From observations 15 and 16 I want to replace these
> observations by a new one with that
> date (=04mar2006), firm is unchanged at 1, value is summed
> (=92+149=241), child and
> height stay unchanged.
>
> --------------------------------------------------------
> obs hh time firm value child height
> 13. 72 08feb2006 4 441 1 3
> 14. 72 14feb2006 2 59 1 3
> 15. 72 04mar2006 1 92 1 3
> 16. 72 04mar2006 1 149 1 3
> 17. 72 08mar2006 4 213 1 3
> --------------------------------------------------
> 18. 72 16mar2006 1 453 1 3
> 19. 72 16mar2006 1 92 1 3
> 20. 72 16mar2006 4 259 1 3
> 21. 72 20mar2006 1 79 1 3
> 22. 72 24mar2006 4 295 1 3
> --------------------------------------------------
> 23. 72 30mar2006 4 279 1 3
> 24. 72 04apr2006 1 111 1 3
> 25. 72 24apr2006 1 99 1 3
> --------------------------------------------------------
>
*
* 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/
