[R] Building a stand-alone package

Mark Myatt mark at myatt.demon.co.uk
Fri May 31 11:50:17 CEST 2002

tlorino at vet-alfort.fr writes:
>Dear R users,
>I'm working with Windows 2000 and R -- note that I could maybe work with Linux 
>I would like to know if it is possible to build a stand-alone statistical
>package which:
>-- could be programmed in R language,
>-- could have a nice graphical interface with buttons and menus (this package
>will be distributed to non-statistician people),
>-- could realize statistical operations, like for example simulations with
>Monte-Carlo method.
>If needed, it could be implemented with Java (or Perl?).

This should be possible as there is TCL/TK support in R as well as
libraries for Java and Python that allow access to R. On systems that
support names pipes (i.e. any UNIX derivative) it should be relatively
easy to send and receive stuff from a R process from just about any
language environment. I have a rough little text editor / output pager
application written in TCL/TK that can (just about) do this that has a
customisable user interface (simple menus configures through a test
file) that I played about with a few months ago. The intention was to
have this handle menus, syntax editing with colour coding, output
display and saving and have R with the TCL/TK support library handle
everything else. Pressure of work stopped development ... the point is
that it is not very difficult. On the Win32 front there is a control
that allows access to R from within (e.g.) VB, VC++, VBA, Delphi. I have
not looked at this though.

The real difficulty is, I think, getting the user interface right. This
is easy for a very simple system with a few basic functions but if you
wanted to a more complex system capable of exploiting the power of the
underlying R engine then the user interface could become a jungle of
option boxes and sub-dialogs ... better in that case, maybe, to stick
with R syntax in an editor that provides code completion with a neat
'object inspector' type interface.

Just my tuppence.


Mark Myatt

r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch

More information about the R-help mailing list