This looks like a possible job for the parmest package, downloadable
from SSC. The parmest package creates an output dataset (or
resultsset) of estimation results, with one observation per parameter
(or per parameter per bygroup) and data on the parameter estimates,
confidence limits, Pvalues or other parameter attributes requested by
the user. This output dataset can then be input into a graphics command,
such as the SSC package eclplot, which
produces confidence interval plots. More about parmest and other
resultssetrelated commands can be found on my website (see my signature
below).
An alternative is to use Benn Jann's estout command, downloadable from
SSC, which can produce spreadsheets of estimation results, which can
also be loaded back into Stata to produce plots.
I hope this helps.
Roger
Roger Newson
Lecturer in Medical Statistics
Respiratory Epidemiology and Public Health Group
National Heart and Lung Institute
Imperial College London
Royal Brompton campus
Room 33, Emmanuel Kaye Building
1B Manresa Road
London SW3 6LR
Tel: +44 (0)20 7352 8121 ext 3381
Fax: +44 (0)20 7351 8322
Email: r.newson@imperial.ac.uk
www.imperial.ac.uk/nhli/r.newson/
Opinions expressed are those of the author, not of the institution.
Original Message
From: ownerstatalist@hsphsun2.harvard.edu
[mailto:ownerstatalist@hsphsun2.harvard.edu] On Behalf Of Sara Mottram
Sent: 08 August 2006 17:02
To: statalist@hsphsun2.harvard.edu
Subject: st: Plotting stored estimates
Apologises if this is a basic question.
I have a binary variable kcons which indicates whether or not the person
has consulted their doctor about knee pain and another variable age_cat,
indicating the age group to which the person belongs.
I have calculated the mean of kcons within each level of age_cat using
mean with the over(), option to get the probability of consulting
with knee pain in each age group. I have then saved these estimates
using estimates store.
I would now like to produce a scatter plot the proportion of people who
consult in each category against age_cat to assess whether there is a
trend.
This sounds like it should be straightforward but I can't work out from
the manuals (I have looked under _estimates in the programming manual)
and I can't find anything in the FAQs about how to do this. If I try
scatter _est age_cat, I get a flat plot. I could of course produce a
plot in Excel by typing in the results of mean, but I feel that this
is probably an important feature that I should learn to use properly?
I have included my code and out put below.
Sara
. mean kcons, over(age_cat)
Mean estimation Number of obs = 742
_subpop_1: age_cat = 5054
_subpop_2: age_cat = 5559
_subpop_3: age_cat = 6064
_subpop_4: age_cat = 6569
_subpop_5: age_cat = 7074
_subpop_6: age_cat = 7579
_subpop_7: age_cat = 80+

Over  Mean Std. Err. [95% Conf. Interval]
+
kcons_p18m 
_subpop_1  .2911392 .0514379 .1901578 .3921207
_subpop_2  .2205882 .0356868 .1505289 .2906475
_subpop_3  .221374 .036413 .1498892 .2928589
_subpop_4  .3164557 .0371185 .2435858 .3893256
_subpop_5  .2605042 .0404049 .1811825 .3398259
_subpop_6  .3766234 .0555805 .2675095 .4857373
_subpop_7  .4047619 .0766573 .2542706 .5552532

. estimates store propnknee
. plot _est age_cat
2 +
e 
s 
a 
m 
p 
l 
e 
( 
) 

f  * * * * * *
*
r 
o 
m 

e 
s 
t 
i 
0 +
++
1 age_cat
7
*
* 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/
*
* 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/