Vandermonde matrix

From Wikipedia, the free encyclopedia

In linear algebra, a Vandermonde matrix, named after Alexandre-Théophile Vandermonde, is a matrix with the monomial terms of a geometric progression in each row, i.e., an m × n matrix

V=\begin{bmatrix} 1 & \alpha_1 & \alpha_1^2 & \dots & \alpha_1^{n-1}\\ 1 & \alpha_2 & \alpha_2^2 & \dots & \alpha_2^{n-1}\\ 1 & \alpha_3 & \alpha_3^2 & \dots & \alpha_3^{n-1}\\ \vdots & \vdots & \vdots & \ddots &\vdots \\ 1 & \alpha_m & \alpha_m^2 & \dots & \alpha_m^{n-1}\\ \end{bmatrix}

or

V_{i,j} = \alpha_i^{j-1}

for all indices i and j. (Some authors use the transpose of the above matrix.)

The determinant of a square Vandermonde matrix (so m=n) can be expressed as:

\det(V) = \prod_{1\le i<j\le n} (\alpha_j-\alpha_i).

The above determinant is sometimes called the discriminant, although many authors, including this encyclopedia, refer to the discriminant as the square of this determinant.

Using the Leibniz formula

\det(V) = \sum_{\sigma \in S_n} \sgn(\sigma) \, \alpha_1^{\sigma(1)-1} \cdots \alpha_n^{\sigma(n)-1},

for the determinant, we can rewrite this formula as

\prod_{1\le i<j\le n} (\alpha_j-\alpha_i) = \sum_{\sigma \in S_n} \sgn(\sigma) \, \alpha_1^{\sigma(1)-1} \cdots \alpha_n^{\sigma(n)-1},

where Sn denotes the set of permutations of {1, 2, ..., n}, and sgn(σ) denotes the signature of the permutation σ.

If m≤n, then the matrix V has maximum rank (m) if and only if all αi are distinct.

When two or more αi are equal, the corresponding polynomial interpolation problem (see below) is ill-posed. In that case one may use a generalization called confluent Vandermonde matrices, which makes the matrix positive definite while retaining most properties. If αi = αi + 1 = ... = αi+k and αi ≠ αi − 1, then the (i + k)th row is given by

V_{i+k,j} = \begin{cases} 0, & \mbox{if } j \le k; \\ \frac{(j-1)!}{(j-k-1)!} \alpha_i^{j-k-1}, & \mbox{if } j > k. \end{cases}

The above formula for confluent Vandermonde matrices can be readily derived by letting two parameters αi and αj go arbitrarily close to each other. The difference vector between the rows corresponding to αi and αj scaled to a constant yields the above equation (for k = 1). Similarly, the cases k > 1 are obtained by higher order differences. Consequently, the confluent rows are derivatives of the original Vandermonde row.

[edit] Applications

These matrices are useful in polynomial interpolation, since solving the system of linear equations Vu = y for u with V the n × n Vandermonde matrix is equivalent to finding the coefficients uj of the polynomial

P(x)=\sum_{j=0}^{n-1} u_j x^j

of degree ≤ n−1 which has the values yi at αi.

The Vandermonde determinant plays a central role in the Frobenius formula, which gives the character of conjugacy classes of representations of the symmetric group.

When the values αk range over powers of a finite field, then the determinant is more commonly known as the Moore determinant, which has a number of interesting properties.

Confluent Vandermonde matrices are used in Hermite interpolation.

A commonly known special Vandermonde matrix is the discrete Fourier transform matrix.

The Vandermonde matrix diagonalizes a companion matrix.

[edit] See also

[edit] References

  • Roger A. Horn and Charles R. Johnson, Topics in matrix analysis, (1991) Cambridge University Press. See Section 6.1.
  • William Fulton and Joe Harris, Representation Theory, A First Course (1991) Springer Verlag New York, ISBN 978-0-387-97495-8 Chapter 4 reviews the representation theory of symmetric groups, including the role of the Vandermonde determinant.
  • David Goss, Basic Structures of Function Field Arithmetic (1996) Springer Verlag New York, ISBN 3-540-63541-6 Chapter 1 reviews the Moore determinant