This preview shows pages 1–3. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: COMPUTING THE NORM OF A MATRIX KEITH CONRAD 1. Norms on Vector Spaces Let V be a vector space over R . A norm on V is a function  : V R satisfying three properties: 1)  v  0, with equality if and only if v = 0, 2)  v + w   v  +  w  for v,w V , 3)  v  =   v  for R , v V . The same definition applies to a complex vector space. From a norm we get a metric on V by d ( v,w ) =  v w  . The standard norm on R n is n X i =1 a i e i = v u u t n X i =1 a 2 i . This gives rise to the Euclidean metric on R n . Another norm on R n is the supnorm: n X i =1 a i e i sup = max i  a i  . This gives rise to the supmetric on R n : d ( a i e i , b i e i ) = max  a i b i  . On C n the standard norm is n X i =1 a i e i = v u u t n X i =1  a i  2 , and the supnorm is defined as on R n . A common way of placing a norm on a real vector space V is via an inner product , which is a pairing ( , ): V V R that is 1) bilinear, 2) symmetric: ( v,w ) = ( w,v ), and 3) positivedefinite: ( v,v ) 0, with equality if and only if v = 0. The standard inner product on R n is n X i =1 a i e i , n X i =1 b i e i ! = n X i =1 a i b i . For an inner product ( , ) on V , a norm can be defined by the formula  v  = p ( v,v ) . That this actually is a norm is a consequence of the CauchySchwarz inequality,  ( v,w )  p ( v,v )( w,w ) =  v  w  , whose proof can be found in most linear algebra books. In 1 2 KEITH CONRAD particular, using the standard inner product on R n we get the classic form of this inequality, as proven by Cauchy: n X i =1 a i b i v u u t n X i =1 a 2 i n X i =1 b 2 i . However, the CauchySchwarz inequality is true for any inner product on a realvector space, not just the standard inner product on R n . The norm on R n that comes from the standard inner product is the standard norm. On the other hand, the supnorm on R n does not arise from an inner product, i.e. there is no inner product whose associated norm is the supnorm. Although the supnorm and the standard norm on R n are not equal, they are each bounded by a constant multiple of the other one: max i  a i  v u u t n X i =1 a 2 i n max i  a i  , i.e.  v  sup  v  n  v  sup . Therefore the metrics these two norms give rise to determine the same notions of convergence: a sequence in R n which is convergent with respect to one of the metrics is also convergent with respect to the other metric. The standard inner product on R n is closely tied to transposition of n by n matrices. For A = ( a ij ) M n ( R ), let A > = ( a ji ) be its transpose. Then for any v,w R n , ( Av,w ) = ( v,A > w ) , where ( , ) is the standard inner product....
View
Full
Document
 Fall '08
 Chandrasekara

Click to edit the document details