R binaries, platform independent and Design of Experiments

Dear Rexpert,

I would like to thank those who spend time  answering my email on the
burning of a CD with all R binary files for Windows and Linux. I have
tried a couple of suggestions but have not been successful. I will pass
these suggestions to our system administrator and I am sure that he can
sort them out.

I hope that the future version of R will be written in Java so that it is
platform independent. This means there will be a single binary file for
each new version of R. At the moment there are files for Windows, Mac OS,
Fedora 1, 2, 3, 4, SUSE 9.1, 9.2, 9.3, etc.

I also hope that the future of version of R include more DOE (Design of
Experiments) modules. R will be more useful if it is not only a language

and environment for statistical computing and graphics, but also for
design of experiments. I am toying with the idea of adding to R some
modules of my Gendex DOE toolkit (http://designcomputing.net/gendex/). I
learn from an Rexpert that the first step for this exercise is to convert
my java code to C++.  As I do not know C++ (and life is short) please let
me know whether there is an alternative way without this conversion and
whether you can actively help me in this exercise.

Regards,

a) The sources are available and really easy to compile on all those
operating systems.

b) You do NOT want to do numerical computations on software available in
Java byte code.

a) The sources are available and really easy to compile on all those
operating systems.

Honestly, I only know how to compile my Java programs. Anyway, we have

been able to work out how to download all files
http://cran.au.r-project.org/. The entire CRAN is 5.4GB. This requires two
DVDs!.

b) You do NOT want to do numerical computations on software available in
Java byte code.

You do not want to do heavy numerical computations with R either. Most
statistical calculation using R requires a fraction of a second and I
cannot see a real difference between say 0.05 second and 0.07 second. NKN.

b) You do NOT want to do numerical computations on software available in
Java byte code.

You do not want to do heavy numerical computations with R either. Most
statistical calculation using R requires a fraction of a second and I
cannot see a real difference between say 0.05 second and 0.07 second. NKN.

On 02/09/05, Nam-Ky Nguyen <nkn at turing.une.edu.au> wrote:

b) You do NOT want to do numerical computations on software available

in

Java byte code.

You do not want to do heavy numerical computations with R either. Most
statistical calculation using R requires a fraction of a second and I
cannot see a real difference between say 0.05 second and 0.07 second.
NKN.

It is my understanding that the problem with Java is that it wasn't
written with serious numerical computation in mind - as far as I know
only in the latest version have Sun started to be address this issue.
The byte code for the java virtual machine has a flawed numerical
model which is not fully compliant with the IEEE754 standard - this
has nothing to do with speed of computation.  Furthermore the integer
model is very restrictive when you want to work on random numbers
using bit-twiddling.