Finding the Adjoint of an Operator on a Hilbert Space

  • Thread starter Thread starter pyf
  • Start date Start date
Click For Summary

Homework Help Overview

The problem involves finding the adjoint of an operator defined on a Hilbert space, specifically an operator \( u \) that maps elements of the space based on an inner product with a fixed vector \( b \) and outputs a scalar multiple of another fixed vector \( a \). The context includes considerations of linearity and the conventions used for inner products.

Discussion Character

  • Conceptual clarification, Assumption checking, Mixed

Approaches and Questions Raised

  • Participants discuss the correctness of the manipulation of inner products and whether the operator is linear or antilinear based on the conventions used. Questions are raised about the implications of these conventions on the definition of the adjoint.

Discussion Status

The discussion is ongoing, with participants exploring different interpretations of linearity in the context of the inner product. Some guidance has been offered regarding the definitions and conventions, but no consensus has been reached on the correct approach to finding the adjoint.

Contextual Notes

There is a noted potential conflict in the conventions used for linearity in the inner product, which may affect the definition of the adjoint operator. Participants are encouraged to verify their definitions and assumptions.

pyf
Messages
5
Reaction score
0

Homework Statement


Let H be a Hilbert space, with a in H and b in H.

Let u be an operator on H with u(x)=<b,x>a

Find the adjoint of u.

Thanks!

Homework Equations


<ux,y>=<x,u*y>


The Attempt at a Solution


<ux,y>=<<b,x>a,y> = <b,x><a,y> = <b<a,y>,x>

So it seems that all I need to do is flip that last inner product round. But that will introduce a complex conjugate, stopping me from getting the equation to the right form (maybe?).
 
Physics news on Phys.org
Is <<b,x>a,y> = <b,x><a,y> really correct?
 
The inner product is linear in its first argument, so I think so.
 
pyf said:
The inner product is linear in its first argument, so I think so.

Ok, I'm used to the convention where it's linear in the second argument and antilinear in the first. Are you sure?
 
I see. Well, this is the form that we've been given. Any ideas?
 
pyf said:
I see. Well, this is the form that we've been given. Any ideas?

Check your definitions again. I have that for a constant c, <x,cy>=c <x,y> and <cx,y>=c* <x,y>. If you take that sort of an inner product the problem works out nicely. If you take the opposite convention then u(cx)=<b,cx>a=(c*)(<b,x>a)=(c*)u(x). That means u ISN'T a linear operator. It's antilinear. The definition of 'adjoint' for antilinear operators needs to be modified. If you REALLY have a backwards convention for the inner product and a forwards convention for the definition of adjoint, then that's messed up.
 

Similar threads

  • · Replies 43 ·
2
Replies
43
Views
5K
  • · Replies 12 ·
Replies
12
Views
1K
  • · Replies 10 ·
Replies
10
Views
2K
  • · Replies 5 ·
Replies
5
Views
3K
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 3 ·
Replies
3
Views
4K
  • · Replies 5 ·
Replies
5
Views
3K
  • · Replies 18 ·
Replies
18
Views
3K
  • · Replies 5 ·
Replies
5
Views
7K