I have got a -reshape- solution of which I do not really think that it
qualifies for "elegant". Here it is:
//unless you have an admission id you need to create one...
gen admid = _n
reshape long v, i(admid) j(admmo)
gen mycond = v == 1 & month == admmo
bys admid (mycond): replace mycond = mycond[_N]
reshape wide v, i(admid) j(admmo)
Regards
Sebastian
On 8/16/07, Nick Cox <[email protected]> wrote:
> Short of a solution using -reshape-, I think
> this will suit:
>
> gen mycond = 0
> forval j = 1/12 {
> replace mycond = 1 if `j' == month & var`j' == 1
> }
>
> where I assume that your -var*- are 1 when true.
>
> Nick
> [email protected]
>
> Jeph Herrin
>
> > I have observations on patient admissions over
> > a 12 month period.
> >
> > For each observation, I have 12 variables, call
> > them var1-var12, which indicate whether a certain
> > condition was met for that patient for months 1-12.
> > So, eg, it may be true for months 1-3 and 7-12.
> >
> > I also have a variable -month-, which indicates
> > the current month of the observed admission. What
> > I want to capture is whether the condition was true
> > in the month of admission.
> >
> > For instance, at first I unthinkingly tried this:
> >
> > gen mycond = var`=month'==1
> >
> > but this of course uses the first observed value
> > of -month-, which happens to be 6, and only checks
> > -var6-.
> >
> > I eventually got there by creating lots of vars, but is
> > there a more direct solution?
> >
>
> *
> * 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/
>
*
* 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/
