Re: st: Encoding string variables: the one var that has no variance encodes differently than the rest

[Nick Cox <njcoxstata@gmail.com>]

What you did was -encode- your variables separately; thus there is no
necessary correspondence between the value labels assigned to each.

Note that -encode- does have a -label()- option that lets you specify
a particular set of value labels to be used. That is official Stata's
way of allowing you to insist on consistent label sets.

In particular, -encode- orders distinct values alphabetically, and in
your last case there was only one distinct value and it was assigned
value 1.

I wrote a program -multencode- (SSC) which encodes a set of variables jointly.

Typing -findit encode- within Stata would have pointed you to this.

Nick
njcoxstata@gmail.com


On 7 May 2013 23:50, David de Jong <davidcdejong@gmail.com> wrote:

> New Stata (12.1) user here, and I'm a first time poster. I hope
> someone might be able to help with something. I've searched the web,
> help files, and archives to no avail.
>
> I tried to encode a series of variables (type str5), with the following:
>
> qui foreach v of var var1-var10 {
>     encode `v', generate(`v'_encoded)
> }
>
> All of the variables seemed to convert as they should (to type long),
> except for the only variable that also happened to have no variance:
> all cases in the original variable had the string "TRUE." Instead of
> converting to a "2" as did all instances of "TRUE" in the other
> variables, it was converted to "1."
>
> The same problem appears when I try it like this:
>
> encode var1, generate(var1_encoded)
>
> I'd be very grateful for any insights into what might be going on.
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/