This preview shows pages 1–6. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Polynomials I Given N n , a monomial in n variables is a function m : R n R defined as m ( x ) := x 1 1 x 2 2 x n n . I The degree of a monomial is defined as deg m := n i =1 i . I A polynomial in n variables is a function p : R n R defined as a finite linear combination of monomials: p := X A c m = X A c x where A N n is a finite set and c R A . I The set of polynomials in n variables { x 1 ,...,x n } will be denoted R [ x 1 ,...,x n ] or, more compactly, R [ x ] . I The degree of a polynomial f is defined as deg f := max A ,c 6 =0 deg m . I R [ x ] will denote the zero polynomial, i.e. ( x ) = 0 x . 18/235 Multipoly Toolbox I Multipoly is a Matlab toolbox for the creation and manipulation of polynomials of one or more variables. I Example: pvar x1 x2 p = 2*x1^4 + 2*x1^3*x2  x1^2*x2^2 + 5*x2^4 I A scalar polynomial of T terms and V variables is stored as a T 1 vector of coefficients, a T V matrix of natural numbers, and a V 1 array of variable names. p .coef = 2 2 1 5 , p .deg = 4 0 3 1 2 2 0 4 , p .var = x1 x2 19/235 Vector Representation I If p is a polynomial of degree d in n variables then there exists a coefficient vector c R l w such that p = c T w where w ( x ) := 1 , x 1 , x 2 , ..., x n , x 2 1 , x 1 x 2 , ..., x 2 n , ..., x d n T l w denotes the length of w . It is easy to verify l w = ( n + d d ) . I Example: p = 2*x1^4 + 2*x1^3*x2  x1^2*x2^2 + 5*x2^4; x = [x1;x2]; w = monomials(x,0:4); c = poly2basis(p,w); p  c*w [c w] 20/235 Gram Matrix Representation I If p is a polynomial of degree 2 d in n variables then there exists a Q S l z l z such that p = z T Qz where z := 1 , x 1 , x 2 , ..., x n , x 2 1 , x 1 x 2 , ..., x 2 n , ..., x d n T The dimension of z is l z = ( n + d d ) . I Equating coefficients of p and z T Qz yields linear equality constraints on the entries of Q I Define q := vec ( Q ) and l w := ( n +2 d 2 d ) . I There exists A R l w l 2 z and c R l w such that p = z T Qz (i.e. p ( x ) z ( x ) T Qz ( x ) ) is equivalent to Aq = c . I There are h := l z ( l z +1) 2 l w linearly independent homogeneous solutions { N i } h i =1 each of which satisfies z T N i z = ( x ) . I Summary : All solutions to p = z T Qz can be expressed as the sum of a particular solution and a homogeneous solution. The set of homogeneous solutions depends on n and d while the particular solution depends on p . 21/235 Gram Matrix Example p = 2*x1^4 + 2*x1^3*x2  x1^2*x2^2 + 5*x2^4; [z,c,A,w] = gramconstraint(p); pc*w Q = full(reshape(A\c,[3 3])); pz*Q*z % Q is a particular solution in vectorized form % Each column of N is a homogenous solution in vectorized form....
View
Full
Document
This note was uploaded on 01/04/2012 for the course CDS 212 taught by Professor Tarraf,d during the Fall '08 term at Caltech.
 Fall '08
 Tarraf,D

Click to edit the document details