[R] ROMA project (R Output MAnager) / Progress report / Call for contributors

[Eric Lecoutre]

Hello useRs,

This email is about the ROMA project, the implementation in pure S of a
reporting system for R. Several month ago, I did then send on R-help a
survey on that (mainly: what features did you expect for a reporting
system). During the following months, I did try different approaches for
the core functionalities, ending now with a set of generic functions of
which I am satisfied. More details about the coding later for those who
are interested.

Basically, my purpose is to provide to R a mechanism to describe
outputs, in term of structure/content. Then, some driver functions would
handle output object to export them in HTML, LaTeX, RTF, raw text, and
so on.
While progressing in that project, I also did realize that it could even
be an intersting alternative to "print." method: each package creator
who write a "print.myclass" function only needs to describe output. If
he could do that once and benefit from HTML, TeX,... this would be of
great advantage.

I am currently seeking for a new job and I dont know if I will have time
later to go on with that project, which appears to be rather big. That's
why I am searching for collaborators. Anyone interested in reporting may
help. See below for a (non exhaustive) list of tasks / competences. I
have to say that in any cases I am interested in keeping involved in
that problematic.

My latest implementation consists in 2 x 1000 lines of highly commented
code. Mainly 1000 lines for the core functions and 1000 lines for the
beginning of the HTML export driver. Each output object consist of a XML
internal representation (using Duncan Temple Lang XML package), which
gives folowing advantages:
- any output structure could be described (even "non-linear")
- any output "brick" could contain things and have properties (that's
XML node / properties)
- we do have a DOM directly
- XML exportation driver is already functional

Some goals which are already achieved:
- describing complex tables (with merged cells, containing footnotes
etc)
- representing output as internal objects (for storing/further
manipulating purposes)
- manipulating those output objects ("+" operator, changing properties)

There is some work for volonteers in numerous areas:
- reading my actual code and helping to improve it
- thinking about required fundamentals "bricks" to describe outputs
(footnote, p-value, comment, are some sample such "bricks")
- implementing functions that describes some standards outputs
- implementing exporting drivers: LaTeX, native PDF, RTF, raw tex
- ...

Please let me know if you are interested in particpating to this
project. If there are enough participants, I will consider using a
shared platform such as freshmeat or other.


Eric


Eric Lecoutre
UCL /  Institut de Statistique
Voie du Roman Pays, 20
1348 Louvain-la-Neuve
Belgium

tel: (+32)(0)10473050
lecoutre at stat.ucl.ac.be
http://www.stat.ucl.ac.be/ISpersonnel/lecoutre

If the statistics are boring, then you've got the wrong numbers. -Edward
Tufte