Co-norm of an invertible linear transformation on R^n

In summary, the conversation discusses the definition of a co-norm for a linear transformation T on the vector space \mathbb{R}^n. It is denoted as m(T) and is defined as the infimum of the set \{|T(x)|:|x|=1\}. The task is to prove that if T is invertible with inverse S, then m(T)=1/\|S\|. The approach to proving this involves considering the relationship between the norm of T and S, and analyzing the set \{|Tx|:|x|=1\}.
  • #1
ianchenmu
10
0

Homework Statement


[itex]|\;|[/itex] is a norm on [itex]\mathbb{R}^n[/itex].
Define the co-norm of the linear transformation [itex]T : \mathbb{R}^n\rightarrow\mathbb{R}^n[/itex] to be
[itex]m(T)=inf\left \{ |T(x)| \;\;\;\; s.t.\;|x|=1 \right \}[/itex]
Prove that if [itex]T[/itex] is invertible with inverse [itex]S[/itex] then [itex]m(T)=\frac{1}{||S||}[/itex].


Homework Equations


n/a


The Attempt at a Solution


I think probably we need to do something with the norm, but I still can't get it... So thank you.
 
Physics news on Phys.org
  • #2
Equalities of the form inf X = A are often proved by showing that the two inequalities inf X ≤ A and inf X ≥ A both hold. Together they imply equality of course. One of these proofs will typically use that inf X a lower bound of X (consider an arbitrary member of X), and the other will typically use that inf is the greatest lower bound of the set.

How is ##\|S\|## defined? Can you prove anything about the relationship between ##\|S\|## and ##\|T\|##?

Edit: I have so far only proved the inequality ##m(T)\leq 1/\|S\|##. The idea that I think looks the most promising for a proof of the equality is to take a closer look at the set ##\{|Tx|:|x|=1\}##. What is its infimum? What is its supremum?
 
Last edited:

FAQ: Co-norm of an invertible linear transformation on R^n

What is the definition of the co-norm of an invertible linear transformation on R^n?

The co-norm of an invertible linear transformation on R^n is the maximum value of the norm of the inverse transformation applied to any vector in the n-dimensional space.

How is the co-norm related to the norm of a linear transformation?

The co-norm is the reciprocal of the norm of the linear transformation. In other words, the co-norm is the maximum value of the norm of the inverse transformation, while the norm is the maximum value of the transformation itself.

Why is the co-norm important in linear algebra?

The co-norm provides a measure of the "stretching" or "squishing" effect of an invertible linear transformation on the n-dimensional space. It is used to determine the stability and convergence of numerical algorithms, and also has applications in optimization and control theory.

How is the co-norm calculated in practice?

In practice, the co-norm can be calculated by finding the norm of the inverse transformation and taking its reciprocal. This involves finding the inverse matrix or using other methods such as Gaussian elimination or QR decomposition.

Can the co-norm be greater than 1?

Yes, the co-norm can be greater than 1. This indicates that the inverse transformation has a greater stretching effect on the n-dimensional space compared to the original transformation. In general, the co-norm can range from 0 to infinity.

Back
Top