logo

📂Abstract Algebra

Elementary Symmetric Polynomials

Definition1

The following polynomial in $n$ variables $x_{1}, \dots, x_{n}$ is called the $k$-th elementary symmetric polynomial.

$$ e_{k}(x_{1}, \dots, x_{n}) := \sum_{\mathclap{1 \le i_{1} < i_{2} < \cdots < i_{k} \le n}} x_{i_{1}} x_{i_{2}} \cdots x_{i_{k}}, \qquad k = 1, \dots, n $$

By convention, we set $e_{0} := 1$, and $e_{k} := 0$ whenever $k < 0$ or $k > n$.

Simple Definition

$e_{k}$ is the expression obtained by choosing $k$ of the $n$ variables, multiplying them together, and then adding up every possible choice without omission. For instance, when there are three variables $x, y, z$, there are only three ways to choose two of them and multiply, namely $xy$, $yz$, $zx$, so $e_{2}$ is as follows.

$$ e_{2}(x, y, z) = xy + yz + zx $$

In the same way, $e_{1}$ chooses one variable at a time, so it is the sum of all the variables $x + y + z$, and $e_{3}$ chooses all three, so it is the product of all the variables $xyz$. Therefore, for $n=3$, each $e_{k}$ is as follows.

$$ \begin{align*} e_{1} &= x + y + z \\ e_{2} &= xy + yz + zx \\ e_{3} &= xyz \end{align*} $$

Explanation

Since there are $\binom{n}{k}$ ways to choose $k$ out of $n$ distinct objects in total, the number of terms in $e_{k}$ is also $\binom{n}{k}$. Moreover, since every term is formed by multiplying $k$ variables, all terms have the same degree $k$, and such a polynomial is called a homogeneous polynomial of degree $k$.

Changing the order of the variables does not alter "all the ways of choosing $k$ variables." In the above $e_{2}(x, y, z) = xy + yz + zx$, if we swap the order of $x$ and $z$, we get $e_{2}(z, y, x) = zy + yx + xz$, which is the same expression with only the order in which the terms are written changed. Translated into the language of the symmetric group $S_{n}$, this means that the following holds for any permutation $\sigma$, and a polynomial with this property is called a symmetric polynomial.

$$ e_{k}(x_{\sigma(1)}, \dots, x_{\sigma(n)}) = e_{k}(x_{1}, \dots, x_{n}) $$

There are countless other symmetric polynomials besides these, and the reason these particular ones are given the name "elementary" is that any symmetric polynomial whatsoever can be expressed uniquely in terms of sums and products of $e_{1}, \dots, e_{n}$. In other words, as long as we are dealing with symmetric polynomials, these $n$ polynomials are all we need.

Example

For $n = 2$, there are the following two.

$$ e_{1} = x_{1} + x_{2}, \qquad e_{2} = x_{1} x_{2} $$

For $n = 3$, there are the following three.

$$ e_{1} = x_{1} + x_{2} + x_{3}, \qquad e_{2} = x_{1}x_{2} + x_{1}x_{3} + x_{2}x_{3}, \qquad e_{3} = x_{1} x_{2} x_{3} $$

These appear directly in the coefficient positions, with only alternating signs attached, when one expands the polynomial whose roots are $x_{1}, \dots, x_{n}$. Expanding the case $n = 3$ directly gives the following. That is, from the viewpoint of the polynomial, $e_{k}$ is the coefficient of the degree-$(k-1)$ term, and it consists of the sums of all products of $k$ roots. This is called Vieta’s formulas.

$$ (x - x_{1})(x - x_{2})(x - x_{3}) = x^{3} - e_{1}x^{2} + e_{2}x - e_{3} $$

Properties

(a) $e_{k}$ is a homogeneous symmetric polynomial of degree $k$ consisting of $\binom{n}{k}$ terms.

(b) The generating function is as follows.

$$ \prod_{i=1}^{n} (1 + x_{i} t) = \sum_{k=0}^{n} e_{k}(x_{1}, \dots, x_{n}) t^{k} $$

(c) The following recurrence relation holds.

$$ e_{k}(x_{1}, \dots, x_{n}) = e_{k}(x_{1}, \dots, x_{n-1}) + x_{n} e_{k-1}(x_{1}, \dots, x_{n-1}) $$

(d) Fundamental theorem of symmetric polynomials: Every symmetric polynomial is uniquely expressed as a polynomial in $e_{1}, \dots, e_{n}$.

See Also