Michael, check your email.
Robert, I wish I could say I was doing research, but the truth is I
was trying (successfully) to solve
http://projecteuler.net/index.php?section=problems&id=215
or maybe http://projecteuler.net/index.php?section=problems&id=244.
Both can be attacked by graph theory means.
Btw, I find project euler problems great "real world" benchmarks for
SAGE, since we know all those problems are solvable within 1 min. in
C. The competitive nature of the site makes people really optimise
their code and all solutions are available in the forums. Thanks for
the c_graph info, I will use it next time I need to deal with big
graphs.
Rado
On May 18, 1:32 pm, Robert Miller <rlmills...@gmail.com> wrote:
> Rado,
>
> First of all, thank you for your improvement!
>
> > I was playing with some big(10^6) graphs and noticed SAGE cannot
> > handle constructing them in good time.
>
> I am wondering, what in particular you are using Sage graphs for?
> Graphs in Sage are currently in a transition period. Some things are
> incredibly fast (using no Python at all), and others are still very
> slow. In particular, you might be able to use the faster "c_graph"
> implementation of Sage graphs:
>
> sage: D={}
> sage: for i in xrange(1,10^3):
> D[i]=[i+1,i-1]
> ....:
> sage: g=Graph(D)
> sage: def test(g):
> for i in xrange(g.order()):
> for j in xrange(g.order()):
> _ = g.has_edge(i,j)
> ....:
> sage: time test(g)
> CPU times: user 3.23 s, sys: 0.01 s, total: 3.24 s
> Wall time: 3.25 s
> sage: g=Graph(D,implementation='c_graph')
> sage: time test(g)
> CPU times: user 1.57 s, sys: 0.00 s, total: 1.57 s
> Wall time: 1.58 s
>
> Note that the test would run vastly faster if the test function were
> written in Cython.
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to
sage-devel-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
