st: plotting model results (re-posted, due to email format problem)

[Michael McCulloch <mm@pinest.org>]

Dear Statalist members,
I wish to plot the results of hazard ratios comparing the effect of 
dropping predictor variables one at a time from the full model. I 
have saved these estimates using -parmby-, and exported them to a 
dataset. Using the command

	. sort estimate
	. list idnum ids estimate if parmseq==1, clean noobs

I have the following data points:
    idnum     idstr   estimate
        6   Smoking       0.65
        3       Age       0.65
        7   Surgery       0.68
        2      Full       0.68
        5      Race       0.68
        8       Rad       0.68
        9     Chemo       0.68
        4       Sex       0.75
        1     Unadj       0.77

My goal is to plot the results with the HR from the "Full" model on 
the far left, with the remaining HRs off to the right in ascending 
order. The following code gets me close to what I want.

	. sort estimate
	. twoway scatter estimate idnum if parmseq==1 & ids!="Unadj", 
mlabel(ids) mlabposition(12)

Can anybody suggest how I can force "Full" to the far left, while 
having the remaining data points to the right, in ascending sort 
order?

Thank you,
Michael McCulloch

*
*   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/