Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: ivreg2 and predict?


From   Nick Cox <[email protected]>
To   [email protected]
Subject   Re: st: ivreg2 and predict?
Date   Thu, 24 Mar 2011 17:56:35 +0000

No thoughts on -ivregress- from me. Nick

On Thu, Mar 24, 2011 at 5:44 PM,  <[email protected]> wrote:
> Many thanks Nick, also for the quick response, and yes unfortunately our
> office intranet prevents us using -ssc- to install .ado files.
>
> Any thoughts on creating an AC-only weighting matrix when using
> 'ivregress'?
>
> Thanks again,
>
> Adrien
>
>
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of Nick Cox
> Sent: 24 March 2011 18:29
> To: [email protected]
> Subject: Re: st: ivreg2 and predict?
>
>
> This sounds like a partial installation. You didn't install a file,
> ivreg2_p.ado. Perhaps you used a browser. It is much better to use
> -ssc- to install.
>
> The contrast is not strange, as -ivregress- is an official command and
> all its associated commands are part of your Stata installation.
>
> Nick
>
> On Thu, Mar 24, 2011 at 5:21 PM,  <[email protected]> wrote:
>
>> I'm trying to run the 'predict' command following the 'ivreg2'
> command.
>> My (simplified) equations are:
>>
>> ivreg2 i (inf_dev = L.inf_dev) if yq<190, gmm2s bw(auto)
>> predict ihat
>>
>> When I run these, I get the following error: "unrecognized command:
>> ivreg2_p" . Is this familiar to anyone?
>>
>> Strangely, the predict command works with 'ivregress', however I'm
>> obliged to use 'ivreg2' as (my 2nd problem) I can't manage to create a
>> weighting matrix to correct for *only* autocorrelation and not
>> heteroskedasticity as well. In other words, my ivregress command for
> the
>> above equation would be:
>>
>> ivregress gmm i (inf_dev = L.inf_dev) if yq<190, wmatrix(hac ba opt)
>>
>> ...but using ivregress in this way produces very different coefficient
>> results to the above 'ivreg2' command (which has the AC-only form I
>> need).
>>
>> Any thoughts on these two questions? (1. why the error with 'predict'
>> and 'ivreg2', and 2. how to create a weighting matrix to correct only
>> for AC not H, using ivregress)

*
*   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/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index