The original variable would seem
to satisfy this criterion. I am
not aware that any of -sort-, -merge-
or -collapse- requires numeric ids.
Alternatively,
gen hhid2 = subinstr(hhid, " ", "_", .)
egen hhid3 = group(hhid2), label
may be of some use.
Nick
[email protected]
Ali Murat
> My concern is that I need to transform the string hhid variable into
> the numeric one If I wantto use Stata commands such as sort, merge,
> collapse that all require a unique id variable for each observation-
> or group. 21blank 2 and blank212 are differen hhid numebers and belong
> to different households. When I use the following command
> destring newhhid, replace ignore(" ")
> hose two hhid numbers turned into the same numeric hhid variable hhid
> that makes impossible to differentiate between them.
Nick Cox
> > Sorry, but what's the problem?
> > Stata did what you asked. If you
> > do in fact believe that the blanks
> > are somehow informative, your original
> > variable is the one to use.
Ali Murat
> > > I am trying to convert a string household
> indentification variable
> > > ,that has leading and embedded blanks, to a numeric one.
> When I used
> > > the destring command, the new numeric household
> identification number
> > > has no longer unique values due to leading and embedded
> blanks in the
> > > orıginal string household identification variable. To be more
> > > precise, I would like to document what I have done so far:
> > >
> > > gen newhhid=hhid
> > >
> > > destring newhhid, replace ignore(" ")
> > >
> > > sort newid
> > >
> > > l hhid newhhid
> > >
> > >
> > >
> > > Obsnumber hhid newhhid
> > >
> > > 161. | 21 1 211 |
> > >
> > > 162. | 21 2 212 |
> > >
> > > 163. | 212 212 |
> > >
> > > 164. | 213 213 |
> > >
> > > 165. | 21 3 213 |
> > >
> > > |------------------------|
> > >
> > > 166. | 214 214 |
> > >
> > > 167. | 21 4 214 |
> > >
> > >
> > >
> > > Please note that the observation 162 and 163 have a
> distinct hhid; the
> > > fist has an embedded blankd, the second has a leading
> blank. However,
> > > after using the destring command, they have same hhid number.
> > >
*
* 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/
