Hi, I was curious, so here is how to use cython with sympy:
get my "cython" branch, e.g. for example by: git clone git://github.com/certik/sympy.git cd sympy git co -b cython origin/cython then run in isympy: In [1]: time divisors(10**8) CPU times: user 10.70 s, sys: 0.00 s, total: 10.70 s Wall time: 10.71 s Out[2]: [1, 2, 4, 5, 8, 10, 16, 20, 25, 32, 40, 50, 64, 80, 100, 125, 128, 160, 200, 2 50, 256, 320, 400, 500, 625, 640, 800, 1000, 1250, 1280, 1600, 2000, 2500, 312 5, 3200, 4000, 5000, 6250, 6400, 8000, 10000, 12500, 15625, 16000, 20000, 2500 0, 31250, 32000, 40000, 50000, 62500, 78125, 80000, 100000, 125000, 156250, 16 0000, 200000, 250000, 312500, 390625, 400000, 500000, 625000, 781250, 800000, 1000000, 1250000, 1562500, 2000000, 2500000, 3125000, 4000000, 5000000, 625000 0, 10000000, 12500000, 20000000, 25000000, 50000000, 100000000] Then compile it with cython: $ ./build.py build_ext --inplace Compiling module sympy.ntheory.divisors ... running build_ext building 'sympy.ntheory.divisors' extension creating build creating build/temp.linux-x86_64-2.6 creating build/temp.linux-x86_64-2.6/home creating build/temp.linux-x86_64-2.6/home/ondrej creating build/temp.linux-x86_64-2.6/home/ondrej/repos creating build/temp.linux-x86_64-2.6/home/ondrej/repos/sympy creating build/temp.linux-x86_64-2.6/home/ondrej/repos/sympy/sympy creating build/temp.linux-x86_64-2.6/home/ondrej/repos/sympy/sympy/ntheory gcc -pthread -fno-strict-aliasing -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -fPIC -I/usr/include/python2.6 -c /home/ondrej/repos/sympy/./sympy/ntheory/divisors.c -o build/temp.linux-x86_64-2.6/home/ondrej/repos/sympy/./sympy/ntheory/divisors.o gcc -pthread -shared -Wl,-O1 -Wl,-Bsymbolic-functions build/temp.linux-x86_64-2.6/home/ondrej/repos/sympy/./sympy/ntheory/divisors.o -o sympy/ntheory/divisors.so and try again: In [1]: time divisors(10**8) CPU times: user 0.45 s, sys: 0.00 s, total: 0.45 s Wall time: 0.45 s Out[2]: [1, 2, 4, 5, 8, 10, 16, 20, 25, 32, 40, 50, 64, 80, 100, 125, 128, 160, 200, 2 50, 256, 320, 400, 500, 625, 640, 800, 1000, 1250, 1280, 1600, 2000, 2500, 312 5, 3200, 4000, 5000, 6250, 6400, 8000, 10000, 12500, 15625, 16000, 20000, 2500 0, 31250, 32000, 40000, 50000, 62500, 78125, 80000, 100000, 125000, 156250, 16 0000, 200000, 250000, 312500, 390625, 400000, 500000, 625000, 781250, 800000, 1000000, 1250000, 1562500, 2000000, 2500000, 3125000, 4000000, 5000000, 625000 0, 10000000, 12500000, 20000000, 25000000, 50000000, 100000000] and you can see, it got 23x faster! Note that this function uses a very inefficient algorithm, but it's a perfect candidate for cythonizing (uses just one for loop with integers) and its a real code in sympy, not something artificial, so that's why I chose it. Note that just compiling it with cython speeds it about 1.5x, and adding this .pxd file next to the .py file speeds it another 15x: ------- $ cat divisors.pxd import cython @cython.locals(i=cython.int) cpdef divisors(int n) ---------- So that's just to show that it will work. Ondrej --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "sympy" group. To post to this group, send email to sympy@googlegroups.com To unsubscribe from this group, send email to sympy+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sympy?hl=en -~----------~----~----~----~------~----~------~--~---
