Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
st: -devnplot- available from SSC
From
Nick Cox <[email protected]>
To
[email protected]
Subject
st: -devnplot- available from SSC
Date
Thu, 3 Feb 2011 21:43:14 +0000
Thanks as usual to Kit Baum, a new package -devnplot- is now available
from SSC including one program of that name. Stata 9 is required.
Use -ssc- to install if interested. Typing
. ssc type devnplot.hlp
gives the full story without a commitment to install.
-devnplot- is for what I call deviation plots, in the absence of any
name known to me. The plots don't really need a name, but the program
does, so that's the way it goes.
The main idea of -devnplot- is very simple and not at all new. Plot each
data point relative to some appropriate mean with a marker symbol and a
spike connecting marker and mean. Typically, it makes sense to order
data from smallest to largest, although naturally other orders can make
sense. With one variable and nothing else said, this is just a minor
variation on a quantile plot and supplies little extra over what is
already easy in Stata.
With a breakdown by one or two categorical variables, the plots become
more interesting, and in effect a variation on dot plots (-dotplot- in
official Stata, -stripplot- from SSC). They can work better than dot
plots in some situations with very unequal group numbers. In various
senses they are most appropriate for anova-type problems, regardless of
whether you are actually using -anova- or -oneway-. The syntax is
certainly designed with a nod to those commands.
The package comes with a .do file which gives a demonstration. There is
even a time series example. Many would recognise plots that they have
seen before. I'd be happy to receive literature references beyond those
given in the help, as well as the usual bug reports and suggestions for
development.
*
* 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/