The very latest version of the -smileplot- package is now available for
download from my website, which can be accessed from within Web-aware Stata
7 by typing
The new version has also been emailed to SSC, and will presumably be
downloadable from there after Kit Baum returns (thanks in advance to Kit).
The -smileplot- package takes, as input, a data set with 1 observation per
estimated parameter and data including the corresponding P-values, carries
out a multiple test procedure on the multiple P-values, and plots the
P-values (on a reverse log scale) on the Y-axis against the corresponding
parameter estimates (or another variable) on the X-axis. The new version
has an improved help file (incorporating suggestions kindly provided by
Nick Cox) and also the option of suppressing the reference line on the
X-axis corresponding to the value of a parameter under the null hypothesis.
The second improvement is useful for plotting P-values against a variable
other than the corresponding parameter estimates. For instance, in a genome
scan, the data points might correspond to genes, the P-values might be from
tests of associations of those genes with a disease, and the X-axis
variable might be the positions of those genes on a chromosome.
I would like to thank Nick Cox for his very helpful advice re the best
possible SMCL style for my .hlp file. Unfortunately, I didn't implement his
tip about using SMCL's West European accented characters (see -[P] smcl-),
because these do not seem to be working, at least in the text fonts
available in my Stata. (I use the Stata 7 executable of 11 June 2002 and
the ado-files of 8 November 2002 under Windows 98/SE.)
Best wishes
Roger
--
Roger Newson
Lecturer in Medical Statistics
Department of Public Health Sciences
King's College London
5th Floor, Capital House
42 Weston Street
London SE1 3QD
United Kingdom
Tel: 020 7848 6648 International +44 20 7848 6648
Fax: 020 7848 6620 International +44 20 7848 6620
or 020 7848 6605 International +44 20 7848 6605
Email: [email protected]
Opinions expressed are those of the author, not the institution.
