R help - Jul 2004 - Creating a minimal package

> From: Gabor Grothendieck
> 
[snip]
> 
> The objective should be that creating a package is as easy as this:
> 
>    f <- function()1; g <- function()2; d <- 3; e <- 4:5
>   
package.skeleton(list=c("f","g","d","e"),
name="AnExample")
>    library(AnExample)
>    f()
>    
> which means that the package needs to be inserted where library will
> find it. It should not be necessary to have an understanding of this.
To the best of my knowledge (which is not saying much, admittedly) no
version of R on any OS would that work.  package.skeleton() only create the
directory tree and populate it with code and template files.  You still need
to package it up from a shell with R CMD build AnExample (or R CMD build
--binary AnExample for Windows, and maybe Mac?).  Then you need the
install.package("AnExample", CRAN=NULL), before you can actually do
library(AnExample).

BTW, for functions/objects that you would like to use over and over, but do
not want to go through the hurdle of packaging, there's a fairly simple way
to achieve similar effect: save all of them in a .rda file, and when you
need them, just attach() the .rda file.

Let me bring up the old phrase:  R is Open Source.  If you want something
bad enough, you can try to make it happen yourself.  A good example is the
scripting patch announced on R-devel by Neil McKay.  Neil had provided
patches to various versions of R for a while.  If the packaging situation
bothers you enough, feel free to contribute.  I'm quite sure DM will be
happy to get help...

Andy

Folks:

Without entering the debate, just a little addendum to Andy's suggestions
and
Duncan's and Gabor's comments about the difficulty of building (simple,
containing no binaries and just a few functions) packages for R.

I have for some time now built such R-code only .rda files. It is a trivial
matter to automatically attach them at startup (through .Rprofile, .First, or
other startup options) so that they are then "transparently" available
for use. I
have also found it to be a simple matter to bundle such customized bunches of
functions for "naive" users into a little UI so that they click on a
desktop icon
and R opens with the necessary functions and data (and maybe even prompts them,
etc.). You could get quite fancy with this with TK-type stuff, but I haven't
found it necessary to do this.

Please note: I know this is only for small stuff, risks clashes because the
function names are global variables, is not for general distribution, etc.
However, in the "Pareto Principle" spirit or doing things simply when
that
suffices, I have found such quick and dirty approaches quite useful. Why bang
your head on a wall if you don't need to?

Cheers,
Bert


"Liaw, Andy" wrote:
> BTW, for functions/objects that you would like to use over and over, but do
> not want to go through the hurdle of packaging, there's a fairly simple
way
> to achieve similar effect: save all of them in a .rda file, and when you
> need them, just attach() the .rda file.
>
> Andy
>
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://www.stat.math.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide!
http://www.R-project.org/posting-guide.html
--

Bert Gunter

Non-Clinical Biostatistics
Genentech
MS: 240B
Phone: 650-467-7374


"The business of the statistician is to catalyze the scientific learning
process."

 -- George E.P. Box