As a model gets more complex, it is less likely that it can
be presented by a graph. y = polynomial(x) is one exception, but
in practice even cubics over a whole range are unlikely to be
useful, in my experience.
But, more to the point, a model gets more complex, you have _more_
to type as you spell out each term as _b[varname] * varname. -predict-
does this all for you. And this is error-prone.
Other than the occasional pedagogic advantage of underlining
what is being done, i.e. plugging estimates into a model
formula, I can't understand your preference here.
Nick
[email protected]
Maarten Buis
> True, though I find it easier to continue to use my approach
> when my model gets more complex.
>
> n j cox wrote:
>
> > Even with Maarten's example, things can be made
> > simpler. Starting as before,
> >
> > sysuse auto, clear
> > gen weight2 = weight^2
> > reg price weight weight2
> >
> > if you
> >
> > predict quad
> > twoway mspline quad weight || scatter price weight
> >
> > you get the data shown as well.
> >
> > Alternatively, if you install the -modeldiag- package
> > from either SSC or the Stata website, then
> >
> > predict quad
> > regplot
> >
> > will give you the same graph. This is because -regplot-
> > defaults to a plot of the observed and predicted response
> > against the first predictor named.
> >
> > Maarten Buis
> >
> > sysuse auto, clear
> > gen weight2 = weight^2
> > reg price weight weight2
> > sum weight
> > twoway function y = _b[_cons]+_b[weight]*x+_b[weight2]*x^2,
> range(1760
> > 4840)
> >
> >
> > Tinna
> >
> > I am using an independent variable in its regular and
> squared forms in
> > an OLS and 2SLS regressions. Can I graph this polynomial
> relationship
> > as indicated by the regression in an easy way?
*
* 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/
