Re: st: Why does nl call all parameters as "_cons"?

[Maarten Buis <maartenlbuis@gmail.com>]

The complete name of the first coefficient is not _cons but
bA_const:_cons, the part before the colon is the equation to which the
coefficient belongs the second part after the colon the variable. This
system works well for almost all estimation commands in Stata, but in
-nl- you pay the price for the high flexibility of that command, and
you will just have to live with the awkward naming conventions that
result.  This means you will later on have to refer to the
coefficients using the entire name, i.e. including the equation,
rather than the short-cut of refering to coefficients by just the
variable name. You can find the complete names by typing after you
estimated your -nl- model: -nl, coeflegend-.

Hope this helps,
Maarten


On Fri, May 24, 2013 at 12:09 AM, Richard Sweeney
<sweeney.richard@gmail.com> wrote:
> I'm using nl to estimate a CES production function. Here is a simple example:
>
> eststo: nl (logy = {bA: const log_tot_cap} + log( (exp({b: const
> xb})*x1^({rho}) + x2^({rho})) / {rho} ) ), ///
> init(rho 2) eps(1e-5) hasconstant(bA_const)
>
> Stata's default output looks as expected. However, when I use esttab
> (or matrix list), the coefficients all have "_cons" added on at the
> end. For example:
>
> ----------------------------
>                       (1)
>                      logy
> ----------------------------
> bA_const
> _cons              -3.354***
>                 (-129.98)
> ----------------------------
> bA_log_tot~p
> _cons               0.589***
>                  (204.35)
> ----------------------------
> b_const
> _cons               0.102***
>                    (5.14)
> ----------------------------
> b_xb
> _cons              -0.114***
>                  (-11.49)
> ----------------------------
> rho
> _cons               0.438***
>                  (162.21)
> ----------------------------
> N                   32704
> ----------------------------
>
> Can someone please tell me what this means? In what sense are these
> parameters constant?
>
> Finally, does anyone know how to suppress this suffix? It makes it
> difficult to refer to specific variables later on when using esttab,
> since they're all called "_cons".
>
> Thank you
>
> Rich
> *
> *   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/



-- 
---------------------------------
Maarten L. Buis
WZB
Reichpietschufer 50
10785 Berlin
Germany

http://www.maartenbuis.nl
---------------------------------
*
*   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/