[R] Writing R packages in an easier way?
Dr. Michael Wolf
m-wolf at muenster.de
Fri Feb 11 08:52:05 CET 2011
Dear R colleagues,
is there an easier way to write R packages for the own use - without RTools and TeX?
With R versions < 2.10.0 it was very easy to write a package. Under the path
with the package name you wrote a description file and built some directories
like "help", "html" and "R" with the special files in a "hand-operated" way. In
the next step I constructed the MD5 check sum file with a program like
"md5summer". Packing the directory with a zipper - and the package work was
finished.
Ok, this was not the official way. And now (R versions > 2.10.0) this method
will not be successful and R doesn't accept those packages. But just for me the
official rules writing a package cost a lot of time if I look to the result I
get. My functions are very specific and have something to do with several
computers I use. So I have not the intention to upload this package to CRAN.
From the point of the costs e. g. I had to learn writing help files in a
TeX-like language. But I'm the typical Word user. My last TeX writings were done
in the 1990s! If I'm changing only a letter in a source file (r-file or help
file) I've to build a new package. Seeing the results in my eyes this is a very
expansive way. It's easier to me to write those files in HTML and to change the
HTML source code. I don't need help files in Rd format.
Does anyone know a easier way? If not just a question to the members of the R
Development Core Team: Could you imagine to open an alternative way of writing
packages? Is it possible to that new versions of R will again accept the build
of packages like in R versions < 2.10.0? Don't misunterstand me: I absolutely
agree that R packages which should become part of CRAN must obey to official
rules like described in the manuals. But I think you can controll this while
uploading a package to CRAN. In my eyes an alternative, easier way writing a R
package would help first-time users and typical MS Word damaged people like me.
Thanks for your help in advance!
Greetings from Muenster/Germany
Dr. Michael Wolf
(E-Mail: m-wolf at muenster.de)
More information about the R-help
mailing list