Best languages for simulation & modeling, DSP, computational physics

Best languages for simulation & modeling, DSP, computational physics

Post by Bill » Thu, 31 Jul 2003 14:32:08



I'm studying for a BS in physics.  I'd like to be able to program better, so
I can do simulation & modeling, digital signal processing, and computational
physics.  I know Matlab pretty well and know a tiny bit of Java, but this
won't go all that far in the job market.  I'd like to learn C/C++.  I guess
what I'd like to know is whether or not Fortran is still worthwhile for
these purposes?  Is it more desirable to know Fortran or Unix? or is there
some other language I should know?

Thanks in advance,
S McGovern

 
 
 

Best languages for simulation & modeling, DSP, computational physics

Post by Igor Kaludje » Thu, 31 Jul 2003 15:21:39



Quote:> I'm studying for a BS in physics.  I'd like to be able to program better,
so
> I can do simulation & modeling, digital signal processing, and
computational
> physics.  I know Matlab pretty well and know a tiny bit of Java, but this
> won't go all that far in the job market.  I'd like to learn C/C++.  I
guess
> what I'd like to know is whether or not Fortran is still worthwhile for
> these purposes?  Is it more desirable to know Fortran or Unix? or is there
> some other language I should know?

I'm working on a FAQ that would address this kind of issues (help wanted!).
In the meanwhile: if you're already familiar with Matlab you might want to
try Simulink (for simulation and DSP).
Wheter or not you need to learn a general purpose language depends on what
kind of simulations you're doing.
OTOH, considering that you're a physics student, knowing a programming
language nowdays is a must.
Arguably Fortran is pretty obsolete (IMHO).
Unix is an operating system, not a programming language.

There are many aspects of simulation modeling, many different methods,
languages, tools, etc.
How to find the right ones for you is not an easy task.
Taking a simulation course at your department would probably be a more
effective way to gain some insight into simulation rather than exploring on
your own.

--
Igor Kaludjer
comp.simulation moderator

 
 
 

Best languages for simulation & modeling, DSP, computational physics

Post by Timo Niemine » Thu, 31 Jul 2003 15:23:22



> I'm studying for a BS in physics.  I'd like to be able to program better, so
> I can do simulation & modeling, digital signal processing, and computational
> physics.  I know Matlab pretty well and know a tiny bit of Java, but this
> won't go all that far in the job market.  I'd like to learn C/C++.  I guess
> what I'd like to know is whether or not Fortran is still worthwhile for
> these purposes?  Is it more desirable to know Fortran or Unix? or is there
> some other language I should know?

As far as computational physics goes, I haven't been able to avoid
FORTRAN. Mainly I use MATLAB/Octave. Other comp types around here use
FORTRAN, C, C++, MATLAB, Python. Be prepared to deal with (and modify or
write) code in whatever language one's co-workers are using.

Best to learn computational methods and good programming skills, which is
mostly independent of language. General computer skills, including
familiarity with unix, are always useful.

--
Timo Nieminen - Home page: http://www.physics.uq.edu.au/people/nieminen/
Shrine to Spirits: http://www.users.bigpond.com/timo_nieminen/spirits.html

 
 
 

Best languages for simulation & modeling, DSP, computational physics

Post by Lawson Englis » Thu, 31 Jul 2003 23:37:15



Quote:> I'm studying for a BS in physics.  I'd like to be able to program better,
so
> I can do simulation & modeling, digital signal processing, and
computational
> physics.  I know Matlab pretty well and know a tiny bit of Java, but this
> won't go all that far in the job market.  I'd like to learn C/C++.  I
guess
> what I'd like to know is whether or not Fortran is still worthwhile for
> these purposes?  Is it more desirable to know Fortran or Unix? or is there
> some other language I should know?

> Thanks in advance,
> S McGovern

You can always learn languages. You need *background* if you want to be more
than just a code-monkey though.

Mathematica is the ultimate formula language/program,  I think.
Check out:

http://www.veryComputer.com/

http://www.veryComputer.com/

The original simulation language was Simula. The successor to Simula was the
original "official" object-oriented language, Smalltalk. The most up-to-date
version of Smalltalk is Squeak, which was a project originally started when
"object oriented programming" and Smalltalk-inventor Alan Kay was a vice
president at Disney Imagineering.

If you want to learn to use the most cutting edge *language* for
simulations, complete with simulation projects designed for kids (hence
"Small"talk), as well as *s, you should check out Squeak:

http://www.veryComputer.com/

You can also join the Squeak community of world-class educators, language
designers, programmers and so on, who can help you get started and give you
ideas for new projects.

http://www.veryComputer.com/

http://www.veryComputer.com/

--
New definition of irony:

'Today's liberal Democrats are like the supporters of the Third Reich of the
'30's and '40's
- they absolutely trusted the government to "make things right". '
-Comment made on the internet by an ardent GW Bush supporter.

 
 
 

Best languages for simulation & modeling, DSP, computational physics

Post by Jay Bra » Fri, 01 Aug 2003 11:12:22



> . . . I'd like to be able to program better, so
> I can do simulation & modeling, . . .

Military simulation, if that interests you, tends to favor C++ these
days for the modeling aspects, with Java increasing in popularity,
especially for GUIs.  Unix variants like Linux and Solaris are the
most popular OSs, with Windows occasionally being chosen.  Fortran
could be encountered in legacy applications.

Very few military models use special-purpose simulation languages.
Some very large legacy simulations are written in Simscript, and some
of these are in active maintenance and development after twenty or
more years.

You can't go wrong if you acquire basic Unix/Linux skills, learn to
write small scripts using bash, Perl, Tcl, and/or Python, and study
object-oriented programming in C++, Java, and/or Python.  This is true
of fields outside of simulation as well.

Jay

 
 
 

Best languages for simulation & modeling, DSP, computational physics

Post by Bill » Sun, 03 Aug 2003 02:13:33


Is Fortran an 'ideal' language? or is it more of a foundation on which
computational sciences have been built?

I'm interested in learning Unix, but I don't I don't see how its useful.
The only benefit I see is stability.  Unless I'm mistaken, stability is
primarily an issue for servers.

Do you, by chance, have an idea of how the future job market will be for
physics with simulation & modeling, digital signal processing, and computer
math, ect.?

Thanks, in advance for any useful info.
Steve McGovern


.


> > I'm studying for a BS in physics.  I'd like to be able to program
better, so
> > I can do simulation & modeling, digital signal processing, and
computational
> > physics.  I know Matlab pretty well and know a tiny bit of Java, but
this
> > won't go all that far in the job market.  I'd like to learn C/C++.  I
guess
> > what I'd like to know is whether or not Fortran is still worthwhile for
> > these purposes?  Is it more desirable to know Fortran or Unix? or is
there
> > some other language I should know?

> As far as computational physics goes, I haven't been able to avoid
> FORTRAN. Mainly I use MATLAB/Octave. Other comp types around here use
> FORTRAN, C, C++, MATLAB, Python. Be prepared to deal with (and modify or
> write) code in whatever language one's co-workers are using.

> Best to learn computational methods and good programming skills, which is
> mostly independent of language. General computer skills, including
> familiarity with unix, are always useful.

> --
> Timo Nieminen - Home page: http://www.physics.uq.edu.au/people/nieminen/
> Shrine to Spirits: http://www.users.bigpond.com/timo_nieminen/spirits.html

 
 
 

Best languages for simulation & modeling, DSP, computational physics

Post by Bill » Sun, 03 Aug 2003 02:24:36


Any major differences in OS's?  Linux, Solaris, Tru64, BSD, Windows, Mac,
BeOS?  Will Tru64 run on an Intel compatible chip?
Thanks
Steve



Quote:> > . . . I'd like to be able to program better, so
> > I can do simulation & modeling, . . .

> Military simulation, if that interests you, tends to favor C++ these
> days for the modeling aspects, with Java increasing in popularity,
> especially for GUIs.  Unix variants like Linux and Solaris are the
> most popular OSs, with Windows occasionally being chosen.  Fortran
> could be encountered in legacy applications.

> Very few military models use special-purpose simulation languages.
> Some very large legacy simulations are written in Simscript, and some
> of these are in active maintenance and development after twenty or
> more years.

> You can't go wrong if you acquire basic Unix/Linux skills, learn to
> write small scripts using bash, Perl, Tcl, and/or Python, and study
> object-oriented programming in C++, Java, and/or Python.  This is true
> of fields outside of simulation as well.

> Jay

 
 
 

Best languages for simulation & modeling, DSP, computational physics

Post by Andrzej Lewandowsk » Sun, 03 Aug 2003 08:55:25




Quote:>Any major differences in OS's?  Linux, Solaris, Tru64, BSD, Windows, Mac,
>BeOS?  Will Tru64 run on an Intel compatible chip?
>Thanks
>Steve

I don't see the reason why you ask this question in the context of
computer simulation...

A.L.

 
 
 

Best languages for simulation & modeling, DSP, computational physics

Post by G.A. Ko » Tue, 12 Aug 2003 04:01:25



> I'm studying for a BS in physics.  I'd like to be able to program better, so
> I can do simulation & modeling, digital signal processing, and computational
> physics.  I know Matlab pretty well and know a tiny bit of Java, but this
> won't go all that far in the job market.  I'd like to learn C/C++.  I guess
> what I'd like to know is whether or not Fortran is still worthwhile for
> these purposes?  Is it more desirable to know Fortran or Unix? or is there
> some other language I should know?

> Thanks in advance,
> S McGovern

- you don't need to know computer languages to do modelling, there are
lots of good programs that let you model in more natural languages.
Relevbant society pages are Society for Computer Simulation and
EUROSIM (probably the best, look at their Argesim page.

For new numerical work in physics as such, Fortran is still best;
otherwise C++ or perhaps Visual Basic are more directly useful,
especially if you want to do Windows programming.

But I suggest you use the Linux operating system, which comes with
excellent free compilers.  Look at freshmeat.com for scientific
programs. OPEN DESIRE is a free simulation program for
differential-equation-type simulation. The most powerful modelling
language is Dymola, but this is not free; there is a free trial
package (dynasim.com). SIMULINK does block-diagram programming, but it
is slow and fairly expensive. Tjere is, zaltogether, an embarrassment
of riches; good luck. GAK

 
 
 

Best languages for simulation & modeling, DSP, computational physics

Post by Meindert de Jo » Wed, 13 Aug 2003 02:19:38


Greetings. FORTRAN is an all purpose simulation langueage. You may find a
paper of ours in my websites: http://www.geocities.com/meindertdejong/ and
http://www.geocities.com/biosanitation/  With a physical-mathematical paper
on simulation of spore dispersal in a vegetation. It took me many years of
work. My newsserver account is about to expire. This week! Please use


>> I'm studying for a BS in physics.  I'd like to be able to program better, so
>> I can do simulation & modeling, digital signal processing, and computational
>> physics.  I know Matlab pretty well and know a tiny bit of Java, but this
>> won't go all that far in the job market.  I'd like to learn C/C++.  I guess
>> what I'd like to know is whether or not Fortran is still worthwhile for
>> these purposes?  Is it more desirable to know Fortran or Unix? or is there
>> some other language I should know?

>> Thanks in advance,
>> S McGovern

>- you don't need to know computer languages to do modelling, there are
>lots of good programs that let you model in more natural languages.
>Relevbant society pages are Society for Computer Simulation and
>EUROSIM (probably the best, look at their Argesim page.

>For new numerical work in physics as such, Fortran is still best;
>otherwise C++ or perhaps Visual Basic are more directly useful,
>especially if you want to do Windows programming.

>But I suggest you use the Linux operating system, which comes with
>excellent free compilers.  Look at freshmeat.com for scientific
>programs. OPEN DESIRE is a free simulation program for
>differential-equation-type simulation. The most powerful modelling
>language is Dymola, but this is not free; there is a free trial
>package (dynasim.com). SIMULINK does block-diagram programming, but it
>is slow and fairly expensive. Tjere is, zaltogether, an embarrassment
>of riches; good luck. GAK

 
 
 

Best languages for simulation & modeling, DSP, computational physics

Post by Stefan M?rk » Wed, 13 Aug 2003 05:28:24


Meindert de Jong schrieb am Mon, 11 Aug 2003 17:19:38 +0000 (UTC):

> Greetings. FORTRAN is an all purpose simulation langueage. You may find a
> paper of ours in my websites: http://www.geocities.com/meindertdejong/ and
> http://www.geocities.com/biosanitation/  With a physical-mathematical paper
> on simulation of spore dispersal in a vegetation. It took me many years of
> work. My newsserver account is about to expire. This week! Please use


>>> I'm studying for a BS in physics.  I'd like to be able to program better, so
>>> I can do simulation & modeling, digital signal processing, and computational
>>> physics.  I know Matlab pretty well and know a tiny bit of Java, but this
>>> won't go all that far in the job market.  I'd like to learn C/C++.  I guess
>>> what I'd like to know is whether or not Fortran is still worthwhile for
>>> these purposes?  Is it more desirable to know Fortran or Unix? or is there
>>> some other language I should know?

>>> Thanks in advance,
>>> S McGovern

>>- you don't need to know computer languages to do modelling, there are
>>lots of good programs that let you model in more natural languages.
>>Relevbant society pages are Society for Computer Simulation and
>>EUROSIM (probably the best, look at their Argesim page.

>>For new numerical work in physics as such, Fortran is still best;
>>otherwise C++ or perhaps Visual Basic are more directly useful,
>>especially if you want to do Windows programming.

>>But I suggest you use the Linux operating system, which comes with
>>excellent free compilers.  Look at freshmeat.com for scientific
>>programs. OPEN DESIRE is a free simulation program for
>>differential-equation-type simulation. The most powerful modelling
>>language is Dymola, but this is not free; there is a free trial
>>package (dynasim.com). SIMULINK does block-diagram programming, but it
>>is slow and fairly expensive. Tjere is, zaltogether, an embarrassment
>>of riches; good luck. GAK

Do you really think that it's a good suggestion to learn FORTRAN these
days? I have to admit that my FORTRAN experience is little. I had to
take a FORTRAN program of a simulation of moving particles in a box and
extend it to my boss's needs. I chose to reprogram it in C++. The
program was quite easy to translate as the syntax was quite readable,
except for the quite cryptic "a.eq.b" saying a equal to b.
My point is that these days C/C++ is becoming more and more like an
industry standard. Knowing how to program in C/C++ is likely to give you
an advantage when you're looking for a job in real life. And I don't
think that there are a lot of things that you can do in FORTRAN but not
in C/C++.

So, FORTRAN knowers, maybe my program was badly done and a bad example
for good FORTRAN; can someone tell me what are the advantages of
FORTRAN? Or would you agree that it is a bit 'outdated'?

ciao
Stefan

--
Fishing for the good times starts with throwing in your lies.
 (Tom Waits)

 
 
 

Best languages for simulation & modeling, DSP, computational physics

Post by Kelvin Hale » Wed, 13 Aug 2003 21:32:54



> Meindert de Jong schrieb am Mon, 11 Aug 2003 17:19:38 +0000 (UTC):
> > Greetings. FORTRAN is an all purpose simulation langueage. You may find a
> > paper of ours in my websites: http://www.geocities.com/meindertdejong/ and
> > http://www.geocities.com/biosanitation/  With a physical-mathematical paper
> > on simulation of spore dispersal in a vegetation. It took me many years of
> > work. My newsserver account is about to expire. This week! Please use



writes:

Quote:> >>> I'm studying for a BS in physics.  I'd like to be able to program better, so
> >>> I can do simulation & modeling, digital signal processing, and computational
> >>> physics.  I know Matlab pretty well and know a tiny bit of Java, but this
> >>> won't go all that far in the job market.  I'd like to learn C/C++.  I guess
> >>> what I'd like to know is whether or not Fortran is still worthwhile for
> >>> these purposes?  Is it more desirable to know Fortran or Unix? or is there
> >>> some other language I should know?

> >>> Thanks in advance,
> >>> S McGovern

> >>- you don't need to know computer languages to do modelling, there are
> >>lots of good programs that let you model in more natural languages.
> >>Relevbant society pages are Society for Computer Simulation and
> >>EUROSIM (probably the best, look at their Argesim page.

> >>For new numerical work in physics as such, Fortran is still best;
> >>otherwise C++ or perhaps Visual Basic are more directly useful,
> >>especially if you want to do Windows programming.

> >>But I suggest you use the Linux operating system, which comes with
> >>excellent free compilers.  Look at freshmeat.com for scientific
> >>programs. OPEN DESIRE is a free simulation program for
> >>differential-equation-type simulation. The most powerful modelling
> >>language is Dymola, but this is not free; there is a free trial
> >>package (dynasim.com). SIMULINK does block-diagram programming, but it
> >>is slow and fairly expensive. Tjere is, zaltogether, an embarrassment
> >>of riches; good luck. GAK

> Do you really think that it's a good suggestion to learn FORTRAN these
> days? I have to admit that my FORTRAN experience is little. I had to
> take a FORTRAN program of a simulation of moving particles in a box and
> extend it to my boss's needs. I chose to reprogram it in C++. The
> program was quite easy to translate as the syntax was quite readable,
> except for the quite cryptic "a.eq.b" saying a equal to b.
> My point is that these days C/C++ is becoming more and more like an
> industry standard. Knowing how to program in C/C++ is likely to give you
> an advantage when you're looking for a job in real life. And I don't
> think that there are a lot of things that you can do in FORTRAN but not
> in C/C++.

> So, FORTRAN knowers, maybe my program was badly done and a bad example
> for good FORTRAN; can someone tell me what are the advantages of
> FORTRAN? Or would you agree that it is a bit 'outdated'?

> ciao
> Stefan

If you want to work as a programmer of computers, then learn C/C++; but if you want
to work in Engineering programming numerical solutions for computer, then FORTRAN is
still most appropriate.

Kelvin B. Hales
Kelvin Hales Associates Limited
Consulting Process Control Engineers

Web: www.khace.com