Re: st: RE: Removing variables with no observation from datasets

[adiallo5@worldbank.org]

Thanks Nick,

DHS are the Macro-International databases on Demographic and Health Surveys.
Many variables in these datasets contains missings observations and are
labelled with NA: I mean:
e.g:
- v224 "Child receives juice, NA"
-tab v224
. no observation

I want to know how to drop these variables.


Cheers.







                                                                                                                                                
                      "Nick Cox"                                                                                                                
                      <n.j.cox@durham.ac.uk>            To:      <Statalist@Hsphsun2.Harvard.Edu>                                               
                      Sent by:                          cc:                                                                                     
                      owner-statalist@hsphsun2.         Subject: st: RE: Removing variables with no observation from datasets                   
                      harvard.edu                                                                                                               
                                                                                                                                                
                                                                                                                                                
                      01/16/2003 09:41 AM                                                                                                       
                      Please respond to                                                                                                         
                      statalist                                                                                                                 
                                                                                                                                                
                                                                                                                                                
                                                                                                                                                
                                                                                                                                                




adiallo5@worldbank.org
>
> I have a list of variables with a certain number of them
> labelled "...some label..., NA" (mostly DHS variables)
> where NA stands for Not Available.
>
> Since it is too hard to manually remove them, I would like
> to know if there exists a certain way to quickly drop them.
>

I am not clear on the details here:
in particular, I don't know what "DHS" means
and whether you mean _values_
when you say _labels_. But I think the
general problem is fairly clear.

There are various ways to approach it.

One which may apply is to decide that any
variable for which all values are identical is, in
essence, uninformative and can thus be
dropped.

Here is code for this. (Warning:
could be slow with very large data sets.)

foreach v of var * {
             sort `v'
             if `v'[1] == `v'[_N] {
                         drop `v'
             }
}

-dropmiss- (STB-60) is another solution,
but it depends on your following
Stata conventions on missing values.

Nick
n.j.cox@durham.ac.uk
*
*   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/