# How to assume the sum of some variables is equal to a constant?

Suppose I have this expression:

$$\frac{a {\left(t_{1} + t_{2} + t_{3} - 1\right)}}{b}$$

This can be simplified to 0 if we assume $t_1 + t_2 + t_3 = 1$.

How can I accomplish this in Sage? I've tried:

var('t1', 't2', 't3', 'a', 'b')
expr1 = ((t1 + t2 + t3 - 1)*a)/b
expr1.full_simplify()

assume(t1 + t2 + t3 == 1)
expr1.full_simplify()


I expected the second call tofull_simplify() to return 0 but it returned the same result as the first call, which is $$\frac{a t_{1} + a t_{2} + a t_{3} - a}{b}$$

edit retag close merge delete

Note that you can work ( in the sense of algebraic geometry) as follows:

sage: R.<t1,t2,t3,a,b> = PolynomialRing(QQ)
sage: Q = R.quotient( t1+t2+t3-1 ).fraction_field()
sage: expr1 = ((t1 + t2 + t3 - 1)*a)/b
sage: expr1
(t1*a + t2*a + t3*a - a)/b
sage: expr1.parent()
Fraction Field of Multivariate Polynomial Ring in t1, t2, t3, a, b over Rational Field
sage: Q( expr1 )
0


Working in the quotient ring modulo the "assumed relations" will always give the reduced expressions without contorsions.

Thanks for the suggestion. Do you have any idea why assume() didn't work?

1

Asking for ?assume we see for instance in

   EXAMPLES:

Assumptions are typically used to ensure certain relations are
evaluated as true that are not true in general.


that the command is not designed to work with simplifications hand in hand, but rather with special, unstructured conditions that make an integral, a square root, etc. be defined, or to have evaluations of special functions like sin and cos. To give an unrelated example where we would also "expect more", let us consider...

sage: var('k'); assume( k, 'integer' );
sage: cos( k*pi ) == (-1)^k
cos(pi*k) == (-1)^k
sage: bool( cos( k*pi ) == (-1)^k )
True


We have to insist to get the whole information.

In your case just insist to get the information: bool( expr1 == 0 ) .

Sort by » oldest newest most voted var('t1', 't2', 't3', 'a', 'b')
expr1 = ((t1 + t2 + t3 - 1)*a)/b
equ = t1 + t2 + t3 == 1
expr1.subs(solve(equ,t1))

more