Re: st: Using cdeco with frequency weights

[Nick Cox <njcoxstata@gmail.com>]

But the information is incomplete. You are asked to explain _where_
you downloaded user-written programs from.

After some messing around, I found that this program can be downloaded by

net install counterfactual, from("http://www.econ.brown.edu/fac/Blaise_Melly/";)

The help for -cdeco_jmp- says that any kind of weight is allowed. But
this is not correct; the code only allows aweights.

I can't speak for whether aweights would do for your purpose, but
there is a mismatch between code and help which you should take up
with the author(s), who may not be members of Statalist.
Alternatively, clone the code and change the program as needed.

Nick

On Fri, Sep 2, 2011 at 12:51 PM, Phumzile Ncube <phumie4@gmail.com> wrote:

> I am sorry, in the last email I mentioned that I am using -cdeco- when
> in fact I am using -cdeco_jmp- .
>
> The rest of the information is still correct.

> On 2 September 2011 13:47, Phumzile Ncube <phumie4@gmail.com> wrote:
>
>> I am estimating gender wage gaps using quantile regression and would
>> like to decompose these gaps into components explained by
>> characteristics and coefficients. In order to do this I downloaded the
>> -rqdeco- command. However, I realised that I can only use this with
>> sample weights and not frequency weights (which are the only ones
>> available on the dataset that I am using). I then decided to use
>> -cdeco-  which, according to the help file, does allow for frequency
>> weights. When I ran -cdeco- I got the following error:
>>
>>
>> fweight not allowed
>> r(101);
>> end of do-file
>> r(101);
>>
>>
>> Has anyone else encountered this problem and if so what did you do to
>> resolve it?

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/