On Mon, Sep 22, 2008 at 6:11 PM, Craig Citro <[EMAIL PROTECTED]> wrote: > > Hi John, > >> That is not in itself slow: >> sage: time pp=pari.primes_up_to_n(10^8) >> CPU times: user 0.69 s, sys: 0.30 s, total: 1.00 s >> Wall time: 1.00 s >> >> However note that >> >> sage: time pp=prime_range(10^8) >> CPU times: user 143.74 s, sys: 1.51 s, total: 145.26 s >> Wall time: 145.93 s >> >> with all the extra time being spent converting pari integers to Sage ones. >> > > Actually, it's not the Pari <--> Sage conversion that's slow: the list > pp is about 6 million entries long, and even for large numbers, we > have: > > sage: n = pari(10**8) > sage: %time for _ in range(6000000): foo = ZZ(n) > CPU times: user 4.40 s, sys: 0.03 s, total: 4.43 s > Wall time: 4.44 s > > I haven't checked into all the details, but I'm pretty sure that > what's actually going so slow is iteration over the Pari list! I just > wrote a few lines of Cython to do this, and here is the timing for the > same problem above, and it seems to work: > > sage: %time ls = my_prime_range(1,10**8) > CPU times: user 1.99 s, sys: 0.89 s, total: 2.88 s > Wall time: 2.89 s > > sage: type(ls[0]) > <type 'sage.rings.integer.Integer'> > sage: ls[-5:] > [99999931, 99999941, 99999959, 99999971, 99999989] > > I'm about to eat dinner, but I'll get this cleaned up (and possibly > start investigating iteration on Pari objects, if I'm right about > that) tonight. It's shameful that we let it be that slow in the first > place! :)
Well, we = "me", and it's a good thing I have no shame! But I'm really glad you guys are fixing this. Regards, William --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
