Beruflich Dokumente
Kultur Dokumente
Aaron Doman
Abstract
This paper will illustrate a simple yet effective method for solving systems of symmet-
ric equations involving rational functions. By manipulating the given equations, one can
construct a polynomial whose roots are the desired solutions.
Suppose we are given two rational functions f and g with the following conditions:
(1) f is a rational function in one variable;
(2) g is a symmetric rational function in n − 1 variables for some integer n ≥ 2; and we wish to
find all distinct complex numbers x1 , x2 , ..., xn such that
By expressing the right hand-side of the ith equation in terms of xi and the symmetric sums of
the variables, we can find a polynomial with roots x1 , x2 , ..., xn . This technique is illustrated in
the following examples.
Since x, y, z are distinct, they must be the three solutions of the equation
c4 = abd + 1
4
d = abc + 1.
Solution: Let s4 = abcd. Multiplying the first equation by a, the second by b, the third by c
and the fourth by d, we obtain the equations
5
a = a + s4
b5 = b + s
4
.
c5 = c + s4
5
d = d + s4 .
15 − 1 − s4 = 0 ⇒ s4 = 0.
Hence, the solutions are (0, −1, i, −i) and permutations (r = 1 and s4 = 0 are equivalent state-
ments).
It may not be too difficult to eliminate one unneeded root, as shown in the previous example.
When there is more than one extra root, however, even the simplest equations rarely give "nice"
answers. The following problem demonstrates the technique for solving symmetric systems of
rational functions.
Using the first two equations, the third becomes s23 + 3s3 = 0. If s3 = 0, then s1 = s2 = 0, so
x = y = z = 0, impossible. Thus s3 = −3 and x, y, z are the roots of
t3 + 3t2 + 3t + 3 = 0 or (t + 1)3 + 2 = 0.
√ √ √
Therefore, the solutions are (−1 − 3 2, −1 − 3 2ω, −1 − 3 2ω 2 ), where ω is a primitive cube root
of unity, and permutations.
The final problem will illustrate a possible extension of the method from solving equations in-
volving rational functions to ones containing other types of functions.