🚀 go-pugleaf

Jaco van Kooten <jvkooten@baan.nl> writes:

>Does anyone know if there exist mathematical and statistical
>routines in Java? A lot of them are available in Fortran and
>C (at netlib), but I need them in Java. 
>O.K. I can rewrite them, but that's a rather boring and time 
>consuming job. (May be it is time to write a C2java compiler.)
>I am also interested in cooperating with someone having the same
>problem. 

At some point I am going to be interested in having simple functions like
random number generation (with a relatively long period), trig functions,
etc... in Java.  I haven't yet programmed in Java or even looked at a
manual or file on how to do it, but I know I am interested in putting
some real-time stat mech calculations on my web page.  I am willing
to contribute any math routines I write to a collaborative project.
I am also interested in simple visualization (i.e. 2d plots).  Something
like this probably exists, but if it doesn't I would be willing to
contribute to that as well. 

Jaco van Kooten wrote:
> 
> Does anyone know if there exist mathematical and statistical
> routines in Java?

There is a class java.lang.Math with some methods.

Maybe that'll help.

-- 
Ciao, Ron.
:-) Keep smilin'

Ronald Aigner
ronalda@eas.iis.fhg.de
ra3@irz.inf.tu-dresden.de
----------------------------------------------------
working at "Frauenhofer Gesellschaft"
project "aFSIM - analoge fault simulator"
----------------------------------------------------
"Nothing is going to happend if you don't show up"

Quoth the Jaco van Kooten (jvkooten@baan.nl):
>Does anyone know if there exist mathematical and statistical
>routines in Java? A lot of them are available in Fortran and
>C (at netlib), but I need them in Java. 
>O.K. I can rewrite them, but that's a rather boring and time 
>consuming job. (May be it is time to write a C2java compiler.)
>I am also interested in cooperating with someone having the same
>problem. 

>Regards,

>Jaco van Kooten
>Baan Research
>Ede, the Netherlands

>jvkooten@baan.nl
I've written a few basic things in "perl"; isn't there some way you can hook
that into Java??

--
M. Edward Borasky

If operating systems had warning labels:
"MVS -- Best if used before December 31, 1999."

Jaco van Kooten wrote:
> 
> Does anyone know if there exist mathematical and statistical
> routines in Java? A lot of them are available in Fortran and
> C (at netlib), but I need them in Java.
> O.K. I can rewrite them, but that's a rather boring and time
> consuming job. (May be it is time to write a C2java compiler.)
> I am also interested in cooperating with someone having the same
> problem.
> 
> Regards,
> 
> Jaco van Kooten
> Baan Research
> Ede, the Netherlands
> 
> jvkooten@baan.nl

You don't need to rewrite them.  You can use Java native methods
as an interface to them.
-- 

Dan Jacobs                   email:    djacobs@BlueSky.net
Tech Tonic Netsystems        voice:    617-273-5825
Burlington, MA

In article <31546CE1.1C18@BlueSky.net>,
Dan Jacobs  <djacobs@BlueSky.net> wrote:
>Jaco van Kooten wrote:
>> 
>> Does anyone know if there exist mathematical and statistical
>> routines in Java? A lot of them are available in Fortran and
>> C (at netlib), but I need them in Java.
>> O.K. I can rewrite them, but that's a rather boring and time
>> consuming job. (May be it is time to write a C2java compiler.)
>> I am also interested in cooperating with someone having the same
>> problem.
>> 
>> Regards,
>> 
>> Jaco van Kooten
>> Baan Research
>> Ede, the Netherlands
>> 
>> jvkooten@baan.nl
>
>You don't need to rewrite them.  You can use Java native methods
>as an interface to them.

Sure, if you don't want to make your code portable.

I can't imagine using Java to do serious number crunching unless
you wanted to be portably, however...
--
Thomas Andrews           Email: thomaso@centerline.com
CenterLine Software      WWW: http://draco.centerline.com:8080/~thomaso/

> Dan Jacobs  <djacobs@BlueSky.net> wrote:
> >Jaco van Kooten wrote:
> >>
> >> Does anyone know if there exist mathematical and statistical
> >> routines in Java? A lot of them are available in Fortran and
> >> C (at netlib), but I need them in Java.
> >> O.K. I can rewrite them, but that's a rather boring and time
> >> consuming job. (May be it is time to write a C2java compiler.)
> >> I am also interested in cooperating with someone having the same
> >> problem.
> >>> >> 
> >> Jaco van Kooten
> >>
> >
> >You don't need to rewrite them.  You can use Java native methods
> >as an interface to them.
> 
> Sure, if you don't want to make your code portable.
> 
> I can't imagine using Java to do serious number crunching unless
> you wanted to be portably, however...
> --
> Thomas Andrews           Email: thomaso@centerline.com
> CenterLine Software      WWW: http://draco.centerline.com:8080/~thomaso/

Indeed, I want my code to make portable. And I hope that in the
future Java becomes faster.

Regards,
Jaco van Kooten 	Email: jvkooten@baan.nl
Baan Research

Thomas Andrews wrote:
> 
> In article <31546CE1.1C18@BlueSky.net>,
> Dan Jacobs  <djacobs@BlueSky.net> wrote:...
> >You don't need to rewrite them.  You can use Java native methods
> >as an interface to them.
> 
> Sure, if you don't want to make your code portable.
> 
> I can't imagine using Java to do serious number crunching unless
> you wanted to be portably, however...
> --
> Thomas Andrews           Email: thomaso@centerline.com
> CenterLine Software      WWW: http://draco.centerline.com:8080/~thomaso/

In general, such native methods are as portable as the libraries to
which they are interfaced.  If you're calling functions such as those
declared in any standard math.h for example, the most you might require
is a smattering of preprocessor directives in the "C" implementation of
the native method, just as you would with *any* "C" code that needed to
build and run on a variety of platforms.

Java doesn't make "C" portable, but if there libraries that support a
common API on a range of platforms, then using native methods to access
those libraries seems preferable IMHO to another, slower wheel.
-- 

Dan Jacobs                   email:    djacobs@BlueSky.net
Tech Tonic Netsystems        voice:    617-273-5825
Burlington, MA

Jaco van Kooten (jvkooten@baan.nl) wrote:
: > Dan Jacobs  <djacobs@BlueSky.net> wrote:
: > >Jaco van Kooten wrote:
: > >>
: > >> Does anyone know if there exist mathematical and statistical
: > >> routines in Java? A lot of them are available in Fortran and
: > >> C (at netlib), but I need them in Java.
: > >> O.K. I can rewrite them, but that's a rather boring and time
: > >> consuming job. (May be it is time to write a C2java compiler.)
: > >> I am also interested in cooperating with someone having the same
: > >> problem.
: > >>> >> 
: > >> Jaco van Kooten
: > >>
: > >
: > >You don't need to rewrite them.  You can use Java native methods
: > >as an interface to them.
: > 
: > Sure, if you don't want to make your code portable.
: > 
: > I can't imagine using Java to do serious number crunching unless
: > you wanted to be portably, however...
: > --
: > Thomas Andrews           Email: thomaso@centerline.com
: > CenterLine Software      WWW: http://draco.centerline.com:8080/~thomaso/

: Indeed, I want my code to make portable. And I hope that in the
: future Java becomes faster.

It is your choice, of course.  However, if you want to get results
now and use a new and clean OO language for scientific computation
the best one is Sather.  Look at http://www.icsi.berkeley.edu/~sather.

Some {but not enough IMHO} of its good ideas seemed to have gone in to
Java (whether by knowledge or by independent invention I don't know). 


: Regards,
: Jaco van Kooten 	Email: jvkooten@baan.nl
: Baan Research

    I recall seeing an announcement in the Numerical Analysts Digest
about LaPack for Java a couple of weeks ago.  Check Netlib for either
the Java implementation or for the NA Digest archives.

Dean Schulze