I was a bit surprised by the following behaviour:
sage: var('c0 c1 c2 x0 x1 x')
(c0, c1, c2, x0, x1, x)
sage: (c0+(x-x0)*(c1+(x-x1)*c2)).collect(x)
c2*x^2 - c2*x*x0 - c2*x*x1 + (c2*x1 - c1)*x0 + c1*x + c0
I assumed that collect
would order my terms into coefficients for powers of x
. At the very least, I'd have expected
c2*x^2 - c2*x*x0 - c2*x*x1 + c1*x + (c2*x1 - c1)*x0 + c0
| x^2 | x x x | 1 |
Although my real goal would have been
c2*x^2 - (c2*x0 - c2*x1 + c1)*x + (c2*x1*x0 - c1*x0 + c0)
I see from the documentation that there is no description at all what collect
does, except returning a symbolic expression. Am I missing the point of that method?
I know I can get at the coefficients as a list using the coeffs
method, but I'd prefer the form as a sum. It seems that even turning that back into a sum, my terms get reordered:
sage: sum([a*x^p for a, p in (c0+(x-x0)*(c1+(x-x1)*c2)).coeffs(x)])
c2*x^2 + (c2*x1 - c1)*x0 - (c2*x0 + c2*x1 - c1)*x + c0
This indicates that the problem might not be in collect
itself, but rather in the way symbolic expressions are stored and printed.
My best solution currently is a manually computed string:
sage: print(' + '.join(['({})*x^{}'.format(a.expand(), p) for a, p in
(c0+(x-x0)*(c1+(x-x1)*c2)).coeffs(x)]))
(c2*x0*x1 - c1*x0 + c0)*x^0 + (-c2*x0 - c2*x1 + c1)*x^1 + (c2)*x^2
Is there a better way to achieve this result?