Re: st: Conditional quantiles graph -- a question for Nick Cox and Maarten Buis

[Nick Cox <njcoxstata@gmail.com>]

Thanks for your interest.

The graphics is just standard -line- calls. The calculations are just
plugging into the conditional quantile function. I don't think I did
it using any canned program, published or unpublished. It was all very
ad hoc. The problem was made easier by the fact that there was only
one predictor, the discharge of the stream, so there is no question of
what precisely you do with the other predictors. This won't match many
other problems, but it reflects the precise scientific problem being
discussed.

I guess it would all be much easier with Stata 12, but the original
work was done nearly 10 years ago.

I'll look for the .do file, but I doubt it differs in spirit from the
description above.

Nick

On Mon, Sep 5, 2011 at 6:13 PM, Gabi Huiber <ghuiber@gmail.com> wrote:
> The latest Stata blog entry presents the merits of poisson over
> regress with a dependent variable in logs. In the comments, Maarten
> recommends a paper Nick wrote. In that paper, there's a graph at
> Figure 3.
>
> Could either Nick or Maarten find the time to post here a code example
> of how one would render such a graph?
>
> Relevant links:
>
> The blog entry:
> http://blog.stata.com/2011/08/22/use-poisson-rather-than-regress-tell-a-friend/
> Nick's paper:
> http://onlinelibrary.wiley.com/doi/10.1002/esp.1523/abstract
>
*
*   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/