Make mod_int signed?

[Volker Braun]

Reposted at the top to increase visiblitiy (see https://groups.google.com/d/msg/sage-devel/5PdRIUic2Es/l0yDYLG_8qIJ)

The sage matrices use unsigned long for the mod_int datatype. This does not fit into a (signed) python int, so whenever it reaches Python (assigning to a Python variable, or passing as an argument to a non-cdef function) it gets converted to a Python arbitrary-precision integer. This wastes quite a bit of time when creating it and gets worse once you start doing arithmetic:

--------------------------------------

ctypedef unsigned long mod_int   # this is how we define it

cpdef caster_slow():

    cdef int i

    foo = None

    for i in range(10000000):

        foo = <mod_int>(1.0)

cpdef caster_fast():

    cdef int i

    foo = None

    for i in range(10000000):

        foo = <int>(1.0)

--------------------------------------

Then we get

sage: timeit('caster_slow()')

5 loops, best of 3: 137 ms per loop

sage: timeit('caster_fast()')

25 loops, best of 3: 34.4 ms per loop

Should we just switch to (signed) long? This wastes one bit (factor of 2) of maximum modulus length, but at least for now we define MAX_MODULUS = 2**23 so this wouldn't be an issue.

[Martin Albrecht]

Sounds like a good idea to me.

Cheers,
Martin

--
name: Martin Albrecht
_pgp: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x8EF0DC99
_otr: 47F43D1A 5D68C36F 468BAEBA 640E8856 D7951CCF
_www: http://martinralbrecht.wordpress.com/
_jab: martinr...@jabber.ccc.de

[Jean-Pierre Flori]

On Tuesday, May 21, 2013 2:47:07 PM UTC+2, Martin Albrecht wrote:

Sounds like a good idea to me.

Sounds good to me too.

[Mike Hansen]

I would just make sure that the unpickling old matrices still works.

--Mike
--Mike

> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-devel+...@googlegroups.com.
> To post to this group, send email to sage-...@googlegroups.com.
> Visit this group at http://groups.google.com/group/sage-devel?hl=en.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

[Volker Braun]

This is now http://trac.sagemath.org/14627  (needs review)