axiom-developer
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Axiom-developer] Tuples


From: Ralf Hemmecke
Subject: Re: [Axiom-developer] Tuples
Date: Sat, 30 Jun 2007 01:00:09 +0200
User-agent: Thunderbird 2.0.0.4 (X11/20070604)

On 06/29/2007 11:36 PM, Franz Lehner wrote:
Hello

here is an algebra question for a change.
Do I understand correctly that there is currently no way to implement
DirectProduct(L: List Group) or DirectProduct(L: Tuple Group)?
That is, one has to manually iterate DirectProduct(G1,G2)
if one wants a direct product of more than two groups?

thanks for your consideration
Franz

Ooops. You seem to be hitting the same problem as me.
If you have a constructor DP which takes a tuple as input, you will run into the problem that it is (nearly) impossible to create actual elements of that thing.

Let's try do define something... Again that will be Aldor, but I hope the problem becomes understandable.

DP(T: Tuple Group): with {
  construct: T -> %
} == add {
  Rep == Array Pointer;
  construct(t: T): % == ...
}

Well the problem is the T here. It is something like (G1, G2, G3) where all groups might be different. Now take (g1, g2, g3) where gi is of type Gi. That thing is of type (G1, G2, G3) but not a Tuple(A) for some domain A, because Tuple constructs homogeneous things. We can only have that the type of (g1, g2, g3) is Cross(G1, G2, G3).

Well, if you use Cross, you must give the types at compile time. But that you cannot do.

However, you have said "iterate DirectProduct(G1,G2)" and that lets me start thinking. I don't claim that I will come up with a good "workaround" but the iteration idea might be something that I could eventually turn into something I like. You have to wait a bit, though. Better listen at the aldor-combinat mailing list. I might forget to post something to axiom-dev.

Ralf




reply via email to

[Prev in Thread] Current Thread [Next in Thread]