I agree with Nick C. I think your best bet is to generate a do-file from
your error dataset. However, if like me you are "scripting
language"-challenged, you can do it using Stata's -file- command. This
little sequence seems to work:
file open myfile using repl_errs.txt, write replace
forvalues n = 1/(# of obs in error dataset) {
local err_var = err_var[`n']
local correct = correct[`n']
local id = id[`n']
file write myfile "replace `err_var' = `correct' in `id'" _n
}
file close myfile
Regards,
Lee
Lee Sieswerda, Epidemiologist
Thunder Bay District Health Unit
999 Balmoral Street
Thunder Bay, Ontario
Canada P7B 6E7
Tel: +1 (807) 625-5957
Fax: +1 (807) 623-2369
[email protected]
www.tbdhu.com
> -----Original Message-----
> From: Nick Cox [SMTP:[email protected]]
> Sent: Monday, January 20, 2003 10:46 AM
> To: [email protected]
> Subject: st: RE: Correcting the original data set
>
> Chau Tak Wai
> >
> > Originally I have a dataset something like this:
> >
> > id x1 x2 ...
> > 1 10 20
> > 2 15 14
> > 3 17 17
> > .. .. ..
> >
> > where id is the identifier of that observation and x1, x2,... are
> > variables we are going to use for analysis. However, this
> > original set
> > of data contain some errors needed to be corrected. Now I
> > have another
> > dataset that contains the error-corrections like this:
> >
> > id err_var correct
> > 1 x1 14
> > 1 x2 18
> > 2 x1 6
> > .. .. ..
> >
> > where err_var contains the variable name that have
> > problems, and correct
> > is the corrected value of that variable. I would like to
> > seek advice on
> > how I can merge the later data set with the first one and
> > replace the
> > wrong values of the first dataset.
> >
>
> This is interesting. My first thought was to -reshape-
> the error-corrections data to -wide- and then -merge, update-.
> but it would seem likely that the -reshape- would
> produce missing values which would then clobber existing
> valid values.
>
> So my second thought is to -reshape- the existing data
> to -long- and then -merge, update- and then -reshape-
> back to -wide-.
>
> Another thought is to read in a text version of
> the error corrections data and convert it line
> by line into a .do file
>
> e.g.
>
> 1 x1 14
>
> becomes
>
> replace x1 = 14 if id == 1
>
> That should be a one-line program in your favourite
> scripting language.
>
> Nick
> [email protected]
>
> *
> * 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/
