[Axiom-developer] Re: iterators and cartesian product.

[Francois Maltey]

Hello, I have subscribed to the two lists open-axiom and fricas.

Waldek Hebisch <address@hidden> writes:

> Concerning more general problem: I think that we also need other approaches.
> More preciesly, it would be good to add beter iteration constructs, which
> make iterating over products easier.  For example, something like:

> [[a, b] for a in 1..10 repeat for b in 1..5] 

> Well, concat version creates intermediate lists and concatenates them.
> Psychologically double iteration is an atomic operation, 
  ---------------
> so it is easier (at least for some folks) to think about. 

All right Waldek ! It's exactly what I think : 
Mathematics writes for (x,y,z) in E X F x G.
And the M... systems allows multiple (not double) iterations 
as an atomic operation. 
Maple isn't perfect for this because we must quote with ' the inner part
    [a $a=0..9] 
but ['10*a+b $a=1..9'$b=0..9] and [''10*a+b $a=1..9'$b=0..9'$c=0..9]
At use, Mupad is almost perfect 
    [a $a=0..9] and [10*a+b $a=1..9$b=0..9] and [10*a+b $a=1..9$b=0..9$c=0..9]
$ is a (right to left)? inner operator.

Of corse parallel iterations must remain as a pretty pure concept of axiom.
The description of these 2 iterations are very short.
It's a right thing to have both in axiom.

About Martin's wish :
---------------------

> What I would like a lot however is something like
 
>         for e in g repeat (a, b) := e ...
> or even for (a, b) in g repeat ...
> given that g returns a Cross.

Other languages allow lists for this uses because they aren't typed.

It's maybe impossible to have parenthesis for cross-product, 
because they are used for functions, but a new constructor as [|...,...|] 
should be a great progress in the interface. 

By example caml uses [1;2;3] for lists and [|1;2;3|] for array.
And mupad allows [a,b] := L when L is a list of 2 terms.
So [a,b] := [b,a] exchanges the values of the variables a and b.
and [a,b,c] := [b,c,a] permutes the 3 values.

The functions makeprod and selectfirst in axiom are too particular 
to this domain. Acces and changes by CC.n might be more fluent.

A new constructor ([|...|] or an other) and acces (by Cross.n) 
don't overload axiom 
because they must exist (do you agree ?),
             are easier to describe (do you agree ??),
         and are connected to the list-syntax (do you agree ???).

I wait with Martin such operators.

F.