the default behaviour of reduce() for ideals

[Thomas Feulner]

Hi,

in the definition of a QuotientRing there is the following assumption

    ASSUMPTION:

    ``I`` has a method ``I.reduce(x)`` returning the normal form
    of elements `x\in R`. In other words, it is required that
    ``I.reduce(x)==I.reduce(y)`` `\iff x-y \in I`, and
    ``x-I.reduce(x) in I``, for all `x,y\in R`.

On the other hand, the default definition of reduce in
sage/rings/ideal.py says
    def reduce(self, f):
        return f       # default

Wouldn't it be better to raise a NotImplementedError?

These are the consequences:

sage: Z16x.<x> = Integers(16)[]
sage: GR.<y> =  Z16x.quotient(x**2 + x+1 )
sage: I = GR.ideal(4)
sage: I.reduce(GR(6))
6 # should be reduced mod 4

another example is
sage: J = Z16x.ideal([x+1, x+1]) # just to avoid that the ideal is identified as a principal ideal
sage: R.<z> = Z16x.quotient(J)
sage: R(x) # should be 15

Best,
Thomas

[Marco Streng]

2012/7/30 Thomas Feulner <thomas....@uni-bayreuth.de>:

> Hi,
>
> in the definition of a QuotientRing there is the following assumption
>
> ASSUMPTION:
>
> ``I`` has a method ``I.reduce(x)`` returning the normal form
> of elements `x\in R`. In other words, it is required that
> ``I.reduce(x)==I.reduce(y)`` `\iff x-y \in I`, and
> ``x-I.reduce(x) in I``, for all `x,y\in R`.
>
> On the other hand, the default definition of reduce in
> sage/rings/ideal.py says
> def reduce(self, f):
> return f # default
>
> Wouldn't it be better to raise a NotImplementedError?

It would be safer, but it sounds like it would break a lot of code.
Right now, reduce does exactly what it says:
"""
Return the reduction of the element of `f` modulo the ideal
`I` (=self). This is an element of `R` that is
equivalent modulo `I` to `f`.
"""

If you don't get any objections, you could change this documentation
and add a deprecation warning to this default implementation. Then
later it can be changed to NotImplementedError.

> These are the consequences:
>
> sage: Z16x.<x> = Integers(16)[]
> sage: GR.<y> = Z16x.quotient(x**2 + x+1 )
> sage: I = GR.ideal(4)
> sage: I.reduce(GR(6))
> 6 # should be reduced mod 4
> another example is
> sage: J = Z16x.ideal([x+1, x+1]) # just to avoid that the ideal is
> identified as a principal ideal
> sage: R.<z> = Z16x.quotient(J)
> sage: R(x) # should be 15

I'm fine with all the outputs above, though they can be improved.
However, I think the following is a bug:

sage: R(x+1) == 0
False

The only correct answer is "True". This goes wrong because the
"ASSUMPTION" that you quotes above is not satisfied. However, I would
trust R to be correct, because I never get to see the assumption, even
when I type

Z16x.quotient??
R??

Suggested fixes:
- add the assumption to the documentation of the quotient method of Z16x, or
- add some checks to the equality tests of QuotientRing, or
- kill the default implementation of reduce as you suggest (but do
deprecation first).

(or some combination of the above)

Best,
Marco

[Marco Streng]

I'm just letting the list know that there is a patch at http://trac.sagemath.org/sage_trac/ticket/13345 deprecating the default implementation of "Ideal_generic.reduce" in favour of NotImplementedError.

Op maandag 30 juli 2012 11:36:06 UTC+2 schreef Marco Streng het volgende:

2012/7/30 Thomas Feulner: