[R] Help with XYPLOT with marginal histograms
gaj.vidmar at mf.uni-lj.si
Thu Feb 12 09:35:45 CET 2009
If you have many data points and/or have to depict several groups of them on
the same graphics, and are -- to gain clarity -- willing to substitute
- points with their <summary> (i.e, convex hull or density contour, plus a
cross or ellipse depicting measures of centre and variability), and
- marginal histograms with either relative frequency polygons or density
just use the package chplot.
It's available at CRAN and extremely easy to use (for R standards) - see the
PDF manual. The added benefit is depicting correlation with the ellipse, and
obtaining a bivariate measure of variability. After you produce the plot
with the chplot function, you can add further elements with the chadd
function. The package also contains one sample (large) datadaset.
A paper (in CMPB, 2005) about is available at
Note that chplot allows you to depict the points themselves, too (and even
skip the convex hull/density contour), but that is basically the oposite of
its main idea, which is to provide clarity with many points (from hundreds
to hundreds of thousands) and/or several groups.
Assist.Prof. Gaj Vidmar, PhD
Institute for Rehabilitation, Republic of Slovenia &
Univ. of Ljubljana, Fac. of Medicine, Inst. of Biomedical Informatics
"Deepayan Sarkar" <deepayan.sarkar at gmail.com> wrote in message
news:eb555e660902111259o1c118567y4e0cc3812dce5d45 at mail.gmail.com...
> On Wed, Feb 11, 2009 at 12:37 PM, John Leonard
> <john.leonard at coe.gatech.edu> wrote:
>> I am trying to replicate the following plot using Lattice:
> Why? lattice is not the right tool for this.
> Try looking at
> with explanations in
> R-help at r-project.org mailing list
> PLEASE do read the posting guide
> and provide commented, minimal, self-contained, reproducible code.
More information about the R-help