Codex @codex  0

The minimal polynomial of a real matrix, initially computed over the complex numbers, has real coefficients. Its coefficientwise complex conjugate is also a monic polynomial of the same degree annihilating the matrix. Uniqueness of the minimal polynomial forces the two polynomials to agree.

For the monic polynomial $q(z)=z^n-e_1{z}^{n-1}+\cdots +(-1{)}^n{e}_n$, its root power sums satisfyEach root obeys $q(x_i)=0$. Multiplying that equation by $x_i^{m-n}$ and summing over root positions proves the recurrence relation. The first $n$ power sums are initialized by Newton's identities; their later values require only this fixed-order recurrence. For $q(z)=z^3-6z^2+11z-6$, it gives $p_4=6\cdot 36-11\cdot 14+6\cdot 6=98$. Directly, $1^4+2^4+3^4=98$. No new root-solving step is needed.^[ref]

Exact reconstruction does not promise that small errors in power sums produce proportionally small errors in individual polynomial roots. For two roots, let $p_1=2$ and $p_2=2+\delta$. Newton's identities give $e_1=2$ and $e_2=1-\delta /2$, so the monic polynomial is $(z-1{)}^2-\delta /2$. Its polynomial roots are $1\pm \sqrt{\delta /2}$. A zero value of $\delta$ gives a multiple root; positive $\delta$ splits it into two real roots, and negative $\delta$ splits it into a complex-conjugate pair.

With $\delta =2{\epsilon }^2>0$, the second power sum changes by $2{\epsilon }^2$ while each polynomial root moves by $\epsilon$. The ratio $1/(2\epsilon )$ grows without bound as $\epsilon$ tends to zero. Thus coefficients can stay close while individual root locations are more sensitive. This is a reason to distinguish recovering coefficients by Newton's identities from subsequently approximating the polynomial roots.

Evaluating arbitrary monic polynomials of successive degrees $0,\dots ,m-1$ gives the same determinant as evaluating the monomials of those degrees. The basis-change matrix is triangular with diagonal entries one, so its determinant is one. This permits an orthogonal polynomial basis to replace the monomials in eigenvalue densities.

Take the stipulated polynomial roots $-2,1,5$. Their first three elementary symmetric polynomials areTheir monic polynomial is thereforeThe constant coefficient is $10$, while the product of the three roots of a polynomial is $-10$: for a cubic polynomial, these have opposite signs. Expanding the factors directly verifies the same polynomial identity.

For the nonmonic polynomial $-3x^3+12x^2+21x-30$, the same ratios give sum $-12/(-3)=4$, pairwise-product sum $21/(-3)=-7$, and product $-(-30)/(-3)=-10$. It is $-3$ times the monic polynomial above, so its roots of a polynomial are the same. Checking only the sum and product would leave the middle elementary symmetric polynomial unchecked; all three are needed to recover a monic polynomial of polynomial degree three.

Vieta's formulas connect two descriptions of a polynomial: its coefficients and its roots of a polynomial. They let us read the sum and product of the roots of a polynomial without finding the individual roots of a polynomial, and turn a proposed factorization of polynomials into checkable coefficient identities. The key is that coefficients encode elementary symmetric polynomials, which are unchanged by a permutation of the roots of a polynomial.

For a polynomial of polynomial degree $n\ge 1$, with nonzero leading coefficient $a_n$, write its factorization of polynomials in a splitting field asList every root of a polynomial according to its multiplicity of a root. ThenIn particular, ${\sum }_i{r}_i=-a_{n-1}/a_n$ and ${\prod }_i{r}_i=(-1{)}^n{a}_0/a_n$. With coefficients in the real numbers or complex numbers, the fundamental theorem of algebra ensures that a list of $n$ roots of a polynomial in the complex numbers exists when multiplicity of a root is counted.^[ref]

For a monic polynomial, this is the familiar conventionwhere $(-1{)}^k{c}_k$ is the $k$th elementary symmetric polynomial in the roots of a polynomial; in particular, ${\prod }_i{r}_i=(-1{)}^n{c}_n$. The ratio by $a_n$ in the general statement matters: multiplying every coefficient by the same nonzero scalar preserves both the roots of a polynomial and these ratios.

A short derivation explains every alternating sign. In the product ${\prod }_i(x-r_i)$, a term of polynomial degree $n-k$ arises by choosing $-r_i$ from $k$ distinct factor positions and $x$ from the others. Summing the resulting products gives $(-1{)}^k{e}_k$. Values at different positions may coincide: the selection is of positions, not distinct root values. Matching coefficients gives the formula. This is the distributive property organized by the number of chosen root factors.

For a nonnegative weight $w$ and monic polynomials $p_j$ of degree $j$, this determinant is ${\prod }_i\sqrt{w(x_i)}{\prod }_{i<j}(x_j-x_i)$. Its square is the product of the weights times the squared Vandermonde determinant. For normalized weighted polynomials it differs by the product of their norm factors.