R help - Jul 2015 - Displaying Compositional Data With 46 Parts

The compositional data have been divided into two data frames: 46 response
variables (the compositional components) and 5 explanatory variables. There
are 209 observations of each. With no experience analyzing large
compositions with so many parts your advice on how to plot and report
results of analyzing these is needed. Matrix plots (scatter, ternary, etc.) of
so many parts would be too small when printed on a page to be readable.

   Haven't found a compositional data water chemistry publication with so
many component parts that could be used as a basis for learning how to work
with such data sets.

   Advice and suggestions needed.

TIA,

Rich

Hi Rich,
I think this is more a technical question for the subject matter experts than
for R-help if I am understanding the question correctly.

That said, there seems to be an R package called compositional and a
corresponding book http://www.springer.com/us/book/9783642368080 that may help. 
If nothing else the names of the author(s) of the package or book  or the
references may give you some leads on some decent papers.

It, probably, would help anyone with some expertise in the area to be able to
look at some of your data.  Have a look at ?dput and perhaps a glance at
http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example
and http://adv-r.had.co.nz/Reproducibility.html for some suggestions on how to
do this.

John Kane
Kingston ON Canada

> -----Original Message-----
> From: rshepard at appl-ecosys.com
> Sent: Fri, 17 Jul 2015 06:16:44 -0700 (PDT)
> To: r-help at r-project.org
> Subject: [R] Displaying Compositional Data With 46 Parts
> 
>    The compositional data have been divided into two data frames: 46
> response
> variables (the compositional components) and 5 explanatory variables.
> There
> are 209 observations of each. With no experience analyzing large
> compositions with so many parts your advice on how to plot and report
> results of analyzing these is needed. Matrix plots (scatter, ternary,
> etc.) of
> so many parts would be too small when printed on a page to be readable.
> 
>    Haven't found a compositional data water chemistry publication with
so
> many component parts that could be used as a basis for learning how to
> work
> with such data sets.
> 
>    Advice and suggestions needed.
> 
> TIA,
> 
> Rich
> 
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
____________________________________________________________
FREE ONLINE PHOTOSHARING - Share your photos online with your friends and
family!
Visit http://www.inbox.com/photosharing to find out more!

On Fri, 17 Jul 2015, John Kane wrote:
> I think this is more a technical question for the subject matter experts
> than for R-help if I am understanding the question correctly.
John,

   I agree completely. Unfortunately, there is no R SIG devoted to CoDA, nor
any other mail list or Web forum that I've been able to find. There is a
CoDA Web site but no active forum.
> That said, there seems to be an R package called compositional and a
> corresponding book http://www.springer.com/us/book/9783642368080 that may
> help. If nothing else the names of the author(s) of the package or book or
> the references may give you some leads on some decent papers.
   In addition to compositions, there's robCompositions and zCompositions.
The book was where I learned how to analyze compositional data. I've
communicated with several of the dozen-or-so statisticians focused on CoDa,
and have a couple of dozen papers, book chapters, and proceedings of the
triennial conferences on compositional data analyses. I've also written a
monograph that demonstrates how CoDA models applied to benthic
macroinvertebrate functional feeding groups can assess water quality and be
used to set standards.

Thanks,

Rich