[Beowulf] Teaching Scientific Computation (looking for the perfect text)

Joe Landman landman at scalableinformatics.com
Tue Nov 20 14:39:29 PST 2007


Nathan Moore wrote:
> 
> 
>      > Nathan,
>      > I'm sure you'll get lots of very experienced responses but if I may:
>      > 1. Book. K&RC is the best book ever, on any subject.
>      > 2. Demographics. It looked to me that engineers were typically
>      > learning and using C (C++, C with Classes, sometimes Java) more than
>      > Fortran. I would have expected similar among physicists, but I
>      > understand that a lot of Fortan is still extant and vital. Also there
>      > is some convergence, ultimately it won't matter much.
> 
> 
>     But for solving a problem (as opposed to learning to get a job
>     programming) what about something like Matlab?  It's procedural, there
>     are compilers (sort of), and it automatically does stuff with matrices
>     in sensible ways.
> 
> 
> No site license for matlab here - I generally have my students couple 

Octave: http://www.gnu.org/software/octave/

After taking students through the joys of programming, I showed them how 
to do masses with springs on Octave.  What a difference.  As Jim Lux 
noted, you spend less time dealing with the vagaries of the language and 
more time helping them articulate a solution (though this particular 
example is bad in that you have many signs you need to correctly and 
carefully account for ... sign errors are a bear in any language)

> gnuplot with some sort of language (perl or fortran depending on how 
> long the job will run), or offer mathematica as an option.

I also like Maxima.

landman at lightning:~$ maxima

Maxima 5.12.0 http://maxima.sourceforge.net
Using Lisp GNU Common Lisp (GCL) GCL 2.6.7 (aka GCL)
Distributed under the GNU Public License. See the file COPYING.
Dedicated to the memory of William Schelter.
This is a development version of Maxima. The function bug_report()
provides bug reporting information.
(%i1) integrate(1/(1+x^2),x,0,inf);
                                       %pi
(%o1)                                 ---
                                        2
(%i2) fortran(%o1);
       %pi/2.0E+0
(%o2)                                done
(%i3)

:)

I used to try to have it help simplify integrals in statistical 
mechanics homework from (owie) 18 years ago.


>     I would certainly eschew any of the fads for "Engineering with Excel"
>     which make my teeth grind when I hear about it.  Every time one of my
>     colleagues creates this incredibly elaborate spreadsheet to calculate
>     receiver performance (gain distribution, intermodulation, etc.) I have
>     to wonder how many hours were spent working around the idiosyncracies
>     of Excel (just to get the plot to look right, if nothing else), when
>     they could have spent that time learning a "real" tool to do the job. 
> 
> 
> Yes, I agree, there is no more asinine task than matrix calculations in 
> excel.  I keep waiting for Microsoft to have competent-looking graphs be 

For fun^h^h^hprofit^h^h^h^h^h^hmasochism I once did a Runge-Kutta orbit 
calculator in Excel.

Yes, you can use it for such things ... but ... why would you want to?


> the default when plotting x&y data.  The new version it even worse than  
> XP excel.  The plots are rendered with some sort of open GL surface so 
> that trend lines now look like giant ropes of licorice. 

Heh...  I still like Gnuplot, as you can programmatically generate input 
decks for it, and have it generate png/jpg/ps/pdf from this ...



-- 
Joseph Landman, Ph.D
Founder and CEO
Scalable Informatics LLC,
email: landman at scalableinformatics.com
web  : http://www.scalableinformatics.com
        http://jackrabbit.scalableinformatics.com
phone: +1 734 786 8423
fax  : +1 866 888 3112
cell : +1 734 612 4615



More information about the Beowulf mailing list