Java Mailing List Archive

http://www.r-help.com/

Home » Home (12/2007) » R Help for Statistical Computing »

Re: [R] Boxplot philosophy {was "Boxplot in R"}

Martin Maechler

2005-07-12

Replies:

>>>>> "Spencer" == Spencer Graves <spencer.graves@(protected)>
>>>>>   on Mon, 11 Jul 2005 19:08:39 -0700 writes:

  Spencer> I'll bite: How does one detect bimodalidty from a boxplot?

One does not,
and that (whole area) was the main reason I mentioned "violin
plots" which do show features such as bimodality:

There's a simple violin plot in the (recommended and hence
always available) package "lattice"

library(lattice)
example(panel.vioplot)

but for the sake of bimodality {and the version of violin plot
that was in the 1998 Am.Stat. paper}

install.packages("vioplot") # if needed
example(vioplot)

Both of these examples use the basic boxplot but *add* to it, so
you can, e.g., clearly discern bimodality.

Martin Maechler

______________________________________________
R-help@(protected)
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
©2008 r-help.com - Jax Systems, LLC, U.S.A.