Hi Kevin,

On 2014-07-27, Kevin Buzzard <[email protected]> wrote:
> [I've just build a degree 6 poly. Now let's build a degree 12 one]
>
> sage: h=expand((g.subs(x+2/x))*x^6)

Let's work without the x^6 factor:

  sage: g
  x^6 + 2*x^3 + x + 1
  sage: g.subs(x+2/x).expand()
  2/x + 1/x^3 + 1/x^6

No idea what is going wrong here. Let's try different ways to
substitute:

  sage: g(x=x+2/x).expand()
  2/x + 1/x^3 + 1/x^6

Same problem. Whithout expanding, it seems to be fine:

  sage: g.subs(x+2/x)
  (((x + 2/x)^3 + 2)*(x + 2/x)^2 + 1)*(x + 2/x) + 1

I guess the above form comes from Horner's scheme.

Now, really strange: If one copy-and-pastes the above expression and
expands it, then all is fine!

  sage: ((((x + 2/x)^3 + 2)*(x + 2/x)^2 + 1)*(x + 2/x) + 1).expand()
  x^6 + 12*x^4 + 2*x^3 + 60*x^2 + 13*x + 26/x + 240/x^2 + 16/x^3 + 192/x^4 + 
64/x^6 + 161

That is all very puzzling to me---and is yet another reason for my creed
that one should use *proper* polynomials and not symbolic expressions
(see below for the underlying types).

> Am I some sort of a victim of some secret property of the letter 'x'?

Not of the letter 'x', but of the fact that the variable x is
pre-defined in Sage as a symbolic expression, which often enough causes
trouble to those who expect to work with polynomials. Note that x is *not*
the generator of a polynomial ring:

  sage: type(x)
  <type 'sage.symbolic.expression.Expression'>

whereas your polynomial g is a proper polynomial (there are various underlying
implementations of polynomials in Sage, for different purposes, and this
one relies on FLINT):

  sage: type(g)
  <type 'sage.rings.polynomial.polynomial_zmod_flint.Polynomial_zmod_flint'>

> I 
> wanted to use x rather than another letter because that would save me from 
> having to define the ring GF(3)[x,1/x] which would have involved some 
> thinking on my part ;-)

Well, the polynomial ring already is there:

  sage: g.parent()
  Univariate Polynomial Ring in x over Finite Field of size 3

and the fraction field will automatically be created when needed. So,
you could do:

  sage: g.parent().inject_variables()
  Defining x                                                              

The preceding line works if you are in an interactive session.
Alternatively (or if you write a program), you could explicitly define
it:

  sage: x = g.parent().gen()
  sage: type(x)
  <type 'sage.rings.polynomial.polynomial_zmod_flint.Polynomial_zmod_flint'>

Now, the substitution works out of the box, and "expansion" of symbolic
expressions is not needed, since we work with proper polynomials and
their quotients.

  sage: g.subs(x+2/x)
  (x^12 + 2*x^9 + x^7 + 2*x^6 + 2*x^5 + x^3 + 1)/x^6

My general recommendation is: Unless you want to do calculus, get used
to define x as the generator of an appropriate polynomial ring, taking
care of the underlying range of coefficients. This is totally easy in an
interactive session:

  sage: P.<x> = QQ[]  # or GF(3)['x'] in your case

The above line both defines P and x, as polynomial ring over the
rationals, and its generator. In a program, you should write

  P = QQ['x']   # or GF(3)['x'] in your case
  x = P.gen()

instead.

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 http://groups.google.com/group/sage-support.
For more options, visit https://groups.google.com/d/optout.

Reply via email to