n-body problem at shootout.alioth.debian.org

Paul McGuire ptmcg at austin.rr._bogus_.com
Fri Oct 6 22:43:25 CEST 2006


"Peter Maas" <peter.maas at somewhere.com> wrote in message 
news:eg6bku$59b$1 at online.de...
>I have noticed that in the language shootout at shootout.alioth.debian.org
> the Python program for the n-body problem is about 50% slower than the 
> Perl
> program. This is an unusual big difference. I tried to make the Python 
> program
> faster but without success. Has anybody an explanation for the difference?
> It's pure math so I expected Perl and Python to have about the same speed.
>
> Peter Maas, Aachen

The advance method is the most fertile place for optimization, since it is 
called approximately n(n-1)/2 times (where n=2E7).  I was able to trim about 
25% from the Python runtime with these changes:

Change:
            distance = (dx**2 + dy**2 + dz**2)**0.5

            b_mass_x_mag = dt * b.mass / distance**3
            b2_mass_x_mag = dt * b2.mass / distance**3
to:
            mag = dt / (dx*dx + dy*dy + dz*dz)**1.5

            b_mass_x_mag = b.mass * mag
            b2_mass_x_mag = b2.mass * mag


And by changing the loop iteration from:

    for i in xrange(len(bodies)) :
        b = bodies[i]
        for j in xrange(i + 1, len(bodies)) :
            b2 = bodies[j]

to:

    remainingBodies = bodies[:]
    for b in bodies[:-1]:
        del remainingBodies[0]
        for b2 in remainingBodies:


-- Paul





More information about the Python-list mailing list