Re: st: graph

[Rufus Browning <browning@sfsu.edu>]

At 11:09 PM 4/6/2003 -0500, Reka Sundaram-Stukel wrote:
> Dear Statalist,
>
> I am a new stata user so would appreciate any help. I estimated an ordered 
> probit model and I would like to create a graph that has probability on 
> the y axis and land on the x axis holding all values (except land) at 
> their mean levels. I know one can do this easily in gauss using the 
> following code. Does any one have any ideas how to do this in STATA. I 
> have tried using predict and plotting it, but it generates the 
> probabilities for each observation.
-prgen- in the excellent SPOSTADO package by Long and Freese will generate 
probabilities based on varying values of a RHS variable with the other RHS 
variables held constant at their means (the default) or at optional other 
values.

findit spostado

will get you to the installable package.

The SPOSTADO package was designed for Stata version 7. In my experience, 
-prtab- in the SPOSTADO package works with version 8, but I have not tried 
-prgen-.

Good luck!

Ref.: J. Scott Long & Jeremy Freese. Regression Models for Categorical 
Dependent Variables Using Stata. College Station, Texas: Stata Press, 2001.

Rufus Browning
Public Research Institute
San Francisco State University



> Thanking you in advance,
>
> Reka
>
> Here is the gauss code:
>
> library pgraph;
>
> /*estimated critical values--first critical vlaues normalized to zero??*/
> c1= -0.28;
> c2= 0.01;
> /*mean/median values for control variables*/
> let z = { means of explanatory variables except land };
>
> /*Estimated coefficients for z variables*/
> let beta = { coeffs of z variables };
> beta_t=0.00652; /*coefficient for land varialbe*/
>
> t=seqa(0.5,1,100);
> beta_x=beta_t*t + beta'z;
>
> probc1=cdfn(c1-beta_x);
> probc3=cdfnc(c2-beta_x);
> probc2=1-probc1-probc3;
>
> xy(t,probc1~probc2~probc3);
>
> saved(
>
>
> Thanking you in advance
> Reka Sundaram-Stukel
>
> Reka Sundaram-Stukel
> Agriculture And Applied Economics
> 320 Taylor Hall
>
> </blockquote></x-html>
*
*   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/