[R] How to search for packages

[john seers (IFR)]

Hi

I think this is a good suggestion. 

And I would like to add the associated problem of deciding between
packages that do the same function which one is "better". Or similarly
packages are often superceded. I find I have to spend a lot of time
learning how to use packages to decide which one is most suitable; or I
spend time with a package to eventually find out it was superceded some
time ago or has been incorporated in another package.

It is also difficult to know how packages fit together (or not).
Functionality often overlaps or is duplicated. I have had some
difficulty with microarray packages - not being sure if they complement
each other or are exclusive to each other. (I know that is more
Bioconductor but the same principle applies in pure R packages, though
perhaps not as pronounced a problem).

Regards

JS

 
---

-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org]
On Behalf Of Monica Pisica
Sent: 04 February 2008 14:34
To: r-help at r-project.org
Subject: [R] How to search for packages


Hi everybody,


I know this might be very off topic and it took me quite a while to up
my courage to post this.... But I remember a thread some time ago about
how we can find the packages we need to do specific tasks in R if we
don't know before hand which ones actually do it. Now all the packages
are listed alphabetically on the web site. Since I am not very advanced
in writing my own functions I relay heavily on work already done and
only when I have no other choice I modify existing functions. Usually my
modifications are only cosmetic.


But sometimes I use lots of time to just read the descriptions of
packages until I decide that maybe one will do close to what I want. I
wonder if there is any way to improve how these packages are displayed
on the site and help with this decision. I wonder if the community as a
whole can come up with some broader categories such as Bayesian, spatial
statistics, bootstrap, vegetation analysis, circular statistics, robust
statistics, etc., and the authors of the package can choose 1 or 2 or
how many categories they think their package fits the most. On the web
page we can have a list of those very broad categories and within each
category we can have in alpha order the packages themselves with their
description and such as it is now. So if I am interested in vegetation
analysis or environmental analysis but I never did it before I go to
that category and see which packages are more geared towards that
particular subject. For example it was by chance alone and some GOOGLE
search that I discovered that the package labdsv has anything to do with
vegetation analysis since first of course I looked at any package which
might have "veg" or "env" in the title.


I also realize that this might mean a lot of work, but R develops so
rapidly that soon I think it will be unmanageable to just peruse the
list of packages and read descriptions in order to choose which package
to install, when you are not familiar with all of them. I hope I didn't
offend the community with this, I would be very sorry since actually I
get lots of help here and I learnt a lot from you. I will remain forever
greatful.



Thanks,

Monica




_________________________________________________________________
[[elided Hotmail spam]]

______________________________________________
R-help at r-project.org mailing list
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.