Jump to content

Sylvester's formula

From Wikipedia, the free encyclopedia

In matrix theory, Sylvester's formula or Sylvester's matrix theorem (named after J. J. Sylvester) or Lagrange−Sylvester interpolation expresses an analytic function f (A) of a matrix A as a polynomial in A, in terms of the eigenvalues and eigenvectors of A.[1][2] It states that[3][4]

f(A)=∑i=1kf(λi) Ai ,

where the λi are the eigenvalues of A, and the matrices

Ai≡∏j=1j≠ik1λi−λj(A−λjI)

are the corresponding Frobenius covariants of A, which are (projection) matrix Lagrange polynomials of A.

Conditions

[edit]

Sylvester's formula applies for any diagonalizable matrix A with k distinct eigenvalues, λ1, ..., λk, and any function f defined on some subset of the complex numbers such that f(A) is well defined. The last condition means that every eigenvalue λi is in the domain of f, and that every eigenvalue λi with multiplicity mi > 1 is in the interior of the domain, with f being (mi - 1) times differentiable at λi.[1]: Def.6.4 

Example

[edit]

Consider the two-by-two matrix:

A=[1342].

This matrix has two eigenvalues, 5 and −2. Its Frobenius covariants are

A1=c1r1=[34][1717]=[37374747]=A+2I5−(−2)A2=c2r2=[17−17][4−3]=[47−37−4737]=A−5I−2−5.

Sylvester's formula then amounts to

f(A)=f(5)A1+f(−2)A2.

For instance, if f is defined by f(x) = x−1, then Sylvester's formula expresses the matrix inverse f(A) = A−1 as

15[37374747]−12[47−37−4737]=[−0.20.30.4−0.1].

Generalization

[edit]

Sylvester's formula is only valid for matrices with no repeated eigenvalues; an extension due to Arthur Buchheim, based on Hermite interpolating polynomials, covers the general case:[5][6]

f(A)=∑i=1s[∑j=0ni−11j!ϕi(j)(λi)(A−λiI)j∏j=1,j≠is(A−λjI)nj],

where ϕi(t):=f(t)/∏j≠i(t−λj)nj.

a concise form of the generalized Sylvester's formula is further given by Hans Schwerdtfeger,[7][4]

f(A)=∑i=1sAi∑j=0ni−1f(j)(λi)j!(A−λiI)j,

where Ai are the corresponding generalized Frobenius covariants of A

Special case

[edit]

If a matrix A is both Hermitian and unitary, then it can only have eigenvalues of ±1, and therefore A=A+−A−, where A+ is the projector onto the subspace with eigenvalue +1, and A− is the projector onto the subspace with eigenvalue −1; By the completeness of the eigenbasis, A++A−=I. Therefore, for any analytic function f,

f(θA)=f(θ)A++f(−θ)A−=f(θ)I+A2+f(−θ)I−A2=f(θ)+f(−θ)2I+f(θ)−f(−θ)2A.

In particular, eiθA=(cos⁡θ)I+(isin⁡θ)A and A=eiπ2(I−A)=e−iπ2(I−A).

See also

[edit]

References

[edit]
  1. 1 2 / Roger A. Horn and Charles R. Johnson (1991), Topics in Matrix Analysis. Cambridge University Press, ISBN 978-0-521-46713-1
  2. ↑ Jon F. Claerbout (1976), Sylvester's matrix theorem, a section of Fundamentals of Geophysical Data Processing. Online version at sepwww.stanford.edu, accessed on 2010-03-14.
  3. ↑ Sylvester, J.J. (1883). "XXXIX. On the equation to the secular inequalities in the planetary theory". The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science. 16 (100): 267–269. doi:10.1080/14786448308627430. ISSN 1941-5982.
  4. 1 2 Higham 2008, §1.10.
  5. ↑ Buchheim, Arthur (August 1886). "An Extension of a Theorem of Professor Sylvester's relating to Matrices". The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science. Fifth Series. 22 (135): 173–174. doi:10.1080/14786448608627914. ISSN 1941-5982.
  6. ↑ Higham 2008, Eq. (1.8).
  7. ↑ Schwerdtfeger, Hans (1938). Les fonctions de matrices: Les fonctions univalentes. I, Volume 1. Paris, France: Hermann.