On Sat, 29 Sep 2018, 10:32 Simon King, <[email protected]> wrote:

> Hi Dima,
>
> On 2018-09-29, Dima Pasechnik <[email protected]> wrote:
> > On Sat, Sep 29, 2018 at 9:58 AM Simon King <[email protected]>
> wrote:
> >>
> >> On 2018-09-29, Simon King <[email protected]> wrote:
> >> > Too bad: When the error occurs and I adjust the pool size then
> >> > afterwards the previously defined objects in gap are gone.
> >>
> >> Additional problem: Even when I increase the memory limit sufficiently,
> >> gap-via-pexpect takes substantially longer than libgap to compute the
> >> double cosets. Why is that? There shouldn't be big traffic in the
> >> interface whan just doing DC = G.DoubleCosetRepsAndSizes(N,N).
> >
> > Just use libGAP, perhaps?
> > (and do the port of the corresponding functionality to libGAP, if you
> have time)
>
> Rewriting my whole p_group_cohomology package to libGAP would be work
> that I'd rather avoid.
>

I believe making libGAP the default GAP interface is the way to go, and
much of the work needed for this conversion is already done...


> Given a group defined in gap-via-pexpect, how to put it into libGAP?
> Slight complication: My groups come with a name. Hence, the string
> representation of the group is not good, libGAP doesn't understand it:
>
>   sage: G = gap.SymmetricGroup(6)
>   sage: G.SetName('"G"')
>   sage: libgap.eval(G)
>   ...
>   ValueError: libGAP: Error, Variable: 'G' must have a value
>

libgap's GAP objects are just thin cython wrappers around GAP's objects in
memory.
So you basically need to send to libgap a complete valid GAP statement
defining G.





> So, how can I make gap produce data that allow to reconstruct the group?
> Is libgap.Group(G.GeneratorsOfGroup()) the way to go? For permutation
> groups probably it is. But for polycyclic groups, as provided by the
> SmallGroups library?
>   sage: G = gap.SmallGroup(720,761)
>   sage: G
>   Group( [ f1, f2, f3, f4, f5, f6, f7 ] )
>   sage: libgap.Group(G.GeneratorsOfGroup())
>   ...
>   ValueError: libGAP: Error, Variable: 'f1' must have a value
>
> Concerning your hint to use the ACE package: After installing
> gap_packages, DC = G.DoubleCosetRepsAndSizes(N,N) takes a very long
> time in gap-via-pexpect (much longer than with libGAP), but so far it
> didn't crash. But it is VERY long. So, rewriting everything to
> libGAP might be the way to go after all. Sigh. Will be tedious work...
>
> Worse: It seems to me that after installing gap_packages, libgap takes
> long as well.
>

ACE is not in gap_packages AFAIK.


> Best regards,
> Simon
>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-support" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To post to this group, send email to [email protected].
> Visit this group at https://groups.google.com/group/sage-support.
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-support" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/sage-support.
For more options, visit https://groups.google.com/d/optout.

Reply via email to