On Oct 16, 2007, at 10:25 AM, John Voight wrote:
>
> Hi Robert,
>
> I can't seem to get primessq to be defined as a cdef long*. I've
> tried
> cdef long primessq[46] = [...]
> cdef long *primessq = [...]
> etc. and they always give errors.
This is because [...] creates a python list, which
Hi Robert,
I can't seem to get primessq to be defined as a cdef long*. I've
tried
cdef long primessq[46] = [...]
cdef long *primessq = [...]
etc. and they always give errors. Cython doesn't allow you to use
macros (#define) I guess?
JV
--~--~-~--~~~---~--~
OK, that makes much more sense.
On Oct 10, 2007, at 1:05 PM, William Stein wrote:
> On 10/10/07, Robert Bradshaw <[EMAIL PROTECTED]> wrote:
>>
>> Only 20% compared to base-10 strings? It seems like there should be
>> more gain than that. Or was it doing something different before?
>
> pari --> s
I filed a related ticket a few month ago:
http://sagetrac.org/sage_trac/ticket/467
david
On Oct 10, 2007, at 4:05 PM, William Stein wrote:
>
> On 10/10/07, Robert Bradshaw <[EMAIL PROTECTED]> wrote:
>>
>> Only 20% compared to base-10 strings? It seems like there should be
>> more gain than tha
On 10/10/07, Robert Bradshaw <[EMAIL PROTECTED]> wrote:
>
> Only 20% compared to base-10 strings? It seems like there should be
> more gain than that. Or was it doing something different before?
pari --> sage hasn't been base-10 strings for years.
Gonzalo wrote optimized code for this in Feb 2006
Only 20% compared to base-10 strings? It seems like there should be
more gain than that. Or was it doing something different before?
- Robert
On Oct 10, 2007, at 10:51 AM, Craig Citro wrote:
> I have a patch to fix that just about ready to go out the door,
> actually. If you want, I could s
I have a patch to fix that just about ready to go out the door, actually. If
you want, I could send it to you and see if it makes your life easier. The
pari->sage speedup is only about 20% (which I'd like to improve), but the
sage->pari is a 5-6X speedup, and gets better with the length of the
inte
Beautiful, thanks! No my unbelievably trivial problem can be solved
extremely fast. There appears to be something else which is
inexplicably eating up time now--maybe coercion between PARI and SAGE
integers in other places...
JV
--~--~-~--~~~---~--~~
To post to
I don't know of any algorithms to find the radical of an integer
quickly (indeed I believe there aren't any) but here are some ideas I
have.
First, if you're dabbling in cython, try making your primessq into a
cdef long*. This should speed things up significantly. If your
calling from oth
On Oct 9, 10:22 pm, John Voight <[EMAIL PROTECTED]> wrote:
> Hello all:
>
> I was hoping you could help me with this unbelievably trivial problem.
>
> Fix an integer B of size about 10 decimal digits, so too big to be an
> unsigned long int but not much bigger.
>
> PROBLEM: Given an integer d of s
I decided on something like this:
-
cimport sage.rings.integer
primessq = [4, 9, 25, 49, 121, 169, 289, 361, 529, 841, 961, 1369,
1681, 1849,
2209, 2809, 3481, 3721, 4489, 5041, 5329, 6241, 6889,
7921, 9409]
len_primes = 25
def int_has_small_square_divisor(sage.rings.integer.In
11 matches
Mail list logo
