On Thursday, 19 July 2012 08:13:04 UTC+8, Rob Beezer wrote:
>
> I'm working (along with a summer research student) to expand the 
> collection of small groups (and their representations) in Sage.  A question 
> about matrix groups, as built by GAP for Sage.
>
> For small, not-very-sophisticated groups, such as a Dihedral group, is it 
> preferable to concoct generators in Sage and use the generic MatrixGroup() 
> constructor or use a GAP routine to build the group?
>
> A simple example:
>
> GAP:  Wrap - "CyclicGroup( IsMatrixGroup, GF(2), 12 )"
>   will build a matrix group of order 12 with matrices of size 12, over a 
> few different types of fields.  This is about as much freedom as GAP allows.
>
> Python:  Mythical - CyclicMatrixGroup(n, R, dimension=None)
>   to build a cyclic group of order n over R, but where "dimension" could 
> default to n, or be sum of invariants, or sum of prime powers, or 1 over 
> the cyclotomic field.  Then an appropriate matrix could be created and sent 
> to the MatrixGroup() (ideally as part of a class providing a sensible repr, 
> etc.)
>
> The second approach allows for a lot more flexibility in providing 
> representations of different dimensions.
>

It looks like a hack. IMHO representations and groups themselves should not 
be mixed in one glass.
Moreover, your approach does not go far enough, as representations of Z_n 
might be quite exotic thing, as soon as n is divisible by 
the characteristic of the field R, or when R isn't a field any more...
(and even in characteristic 0, when your field does not contain the 
splitting field of the group, things are rather nontrivial, IMHO)

Dima

 

>
> My question: is much lost by building the generators and computing the GAP 
> group from those, versus using the built-in GAP named constructions?  
> Again, I'm just interested right now about simpler matrix groups, not the 
> classical groups like GL(), PSL(), etc.
>
> Thanks,
> Rob
>

-- 
-- 
To post to this group, send an email to [email protected]
To unsubscribe from this group, send an email to 
[email protected]
For more options, visit this group at http://groups.google.com/group/sage-devel
URL: http://www.sagemath.org



Reply via email to