Show A = UDU(dagger) can be written as f(A) = Uf(D)U(dagger)

  • Thread starter Thread starter Smazmbazm
  • Start date Start date
Smazmbazm
Messages
45
Reaction score
0

Homework Statement



A diagonalisable square matrix A can be written A = UDU\dagger, where U is a unitary matrix and D is diagonal. Show that nay function of A defined by a power series,

f(A) = f_{0}I + f_{1}A + f_{2}A^{2} + ... + f_{n}A^{n} + ...

can be expressed as f(A) = Uf(D)U\dagger

The Attempt at a Solution



Not sure where to start. I know one can repeat a linear operator to get results such as A^{0} = 1, A^{1} = A, A^{2} = AA but how do you show that for D? Are we simply doing the same? D^{0} = 1, D^{1} = D, D^{2} = DD

So f(A) = U * f_{0}I + f_{1}D + f_{2}D^{2} + ... + f_{n}D^{n} + ... * U\dagger ?

Thanks
 
Physics news on Phys.org
Smazmbazm said:

Homework Statement



A diagonalisable square matrix A can be written A = UDU\dagger, where U is a unitary matrix and D is diagonal. Show that nay function of A defined by a power series,

f(A) = f_{0}I + f_{1}A + f_{2}A^{2} + ... + f_{n}A^{n} + ...

can be expressed as f(A) = Uf(D)U\dagger

The Attempt at a Solution



Not sure where to start. I know one can repeat a linear operator to get results such as A^{0} = 1, A^{1} = A, A^{2} = AA but how do you show that for D? Are we simply doing the same? D^{0} = 1, D^{1} = D, D^{2} = DD

So f(A) = U * f_{0}I + f_{1}D + f_{2}D^{2} + ... + f_{n}D^{n} + ... * U\dagger ?

Thanks
That last line is messed up in some way. If you meant to start the right-hand side with ##U(\dots## and end it with ##\dots)U^\dagger##, then you have just written down the equality you're supposed to prove. But yes, that thing in the middle is ##f(D)##. So it can't be a bad idea to try to use that (the definition of ##f(D)##) to rewrite ##Uf(D)U^\dagger## in some way.
 
Should it look like this,

f(A) = U*f_{0}I*U\dagger + U*f_{1}D*U\dagger + U*f_{2}D^{2}*U\dagger + ... + U*f_{n}D^{n}*U\dagger + ... ?
 
I assumed that you you were trying to write down the equality ##f(A)=Uf(D)U^\dagger##, with ##f(D)=1+f_0 D+\cdots+f_n D^n+\cdots##. You obviously can't assume that the former equality holds, since that's what you want to prove, but you can start
$$Uf(D)U^\dagger=U\left(1+f_0 D+\cdots+f_n D^n+\cdots\right)U^\dagger,$$ to see if you can end up with ##f(A)##. The right-hand side above can be written the way you just wrote it, but I wouldn't use ##*## for multiplication. It's standard to not use any symbol at all, and if you want to use one, it should be ##\circ##, since multiplication of linear operators is just composition of functions.
 
Ok got it. Thanks for your help, Fredrik
 
Prove $$\int\limits_0^{\sqrt2/4}\frac{1}{\sqrt{x-x^2}}\arcsin\sqrt{\frac{(x-1)\left(x-1+x\sqrt{9-16x}\right)}{1-2x}} \, \mathrm dx = \frac{\pi^2}{8}.$$ Let $$I = \int\limits_0^{\sqrt 2 / 4}\frac{1}{\sqrt{x-x^2}}\arcsin\sqrt{\frac{(x-1)\left(x-1+x\sqrt{9-16x}\right)}{1-2x}} \, \mathrm dx. \tag{1}$$ The representation integral of ##\arcsin## is $$\arcsin u = \int\limits_{0}^{1} \frac{\mathrm dt}{\sqrt{1-t^2}}, \qquad 0 \leqslant u \leqslant 1.$$ Plugging identity above into ##(1)## with ##u...
Back
Top