Is the Transpose of the Inverse of a Matrix Its Inverse?

  • Thread starter Thread starter Saladsamurai
  • Start date Start date
  • Tags Tags
    Matrices Work
Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
8 replies · 2K views
Saladsamurai
Messages
3,009
Reaction score
7

Homework Statement



(a) Show that for any invertible matrix A,

[itex](A^{-1})^TA^T=I[/itex] and [itex]A^T(A^{-1})^T=I[/itex]

(b) Deduce that AT is invertible and that its inverse is the transpose of [itex]A^{-1}[/itex]

(c) Deduce also that if A is symmetric then A-1 is also symmetric.


Homework Equations

(AB)T=BTAT



The Attempt at a Solution



(a) If A is invertible,

[itex]AA^{-1}=A^{-1}A=I[/itex]
[itex]\Rightarrow (AA^{-1})^T=I^T[/itex]
[itex]\Rightarrow (AA^{-1})^T=I[/itex]
[itex]\Rightarrow (A^{-1})^TA^T=I[/itex]


Now for part (b) and (c)
 
Physics news on Phys.org
Dick said:
Yes, that works GREAT! Doesn't that make b) pretty easy?

Oh yeah. I guess it kind of takes care of it right? Since this last line [itex](A^{-1})^TA^T=I[/itex] is the definition of an Inverse? That, is: if AB=BA=I then B=A^{-1}

So I also have to show, somehow, that [itex]A^T(A^{-1})^T=I[/itex] as well?
 
Saladsamurai said:
Oh yeah. I guess it kind of takes care of it right? Since this last line [itex](A^{-1})^TA^T=I[/itex] is the definition of an Inverse? That, is: if AB=BA=I then B=A^{-1}

So I also have to show, somehow, that [itex]A^T(A^{-1})^T=I[/itex] as well?

The definition does say 'if AB=BA=I'. So you'd better show both. There's nothing hard about it.
 
Dick said:
The definition does say 'if AB=BA=I'. So you'd better show both. There's nothing hard about it.
If A is invertible, [itex]AA^{-1}=A^{-1}A=I[/itex][itex]\Rightarrow A^{-1}A=I[/itex]
[itex]\Rightarrow (A^{-1}A)^T=I^T[/itex]
[itex]\Rightarrow (A^{-1}A)^T=I[/itex]
[itex]\Rightarrow A^T(A^{-1})^T=I[/itex]

Alright-then :smile:

Now how about part (c). . . Deduce also that if A is symmetric then A-1 is also symmetric.

If A is symmetric, A=AT and if A-1 is symmetric, A-1=(A-1)T

Let me think for a minute here...
 
Remember:

[tex](\mathbf{A}^\mathrm{T})^{-1} = (\mathbf{A}^{-1})^\mathrm{T}[/tex]
 
dirk_mec1 said:
Remember:

[tex](\mathbf{A}^\mathrm{T})^{-1} = (\mathbf{A}^{-1})^\mathrm{T}[/tex]

Is this a property of matrices, or just of exponents in general? Is this just saying that exponents can 'commute'? Sorry if that seems like a stupid question :redface:

:smile:
 
Saladsamurai said:
Is this a property of matrices, or just of exponents in general? Is this just saying that exponents can 'commute'? Sorry if that seems like a stupid question :redface:

:smile:
It's a property of the transpose of matrices I think you need it here.
 
Saladsamurai said:
Is this a property of matrices, or just of exponents in general? Is this just saying that exponents can 'commute'? Sorry if that seems like a stupid question :redface:

:smile:

No, you can't 'commute' everything that's written as a superscript. But you have already shown that [tex]A^T(A^{-1})^T=I[/tex]. That means that the second matrix is the inverse of the first. The inverse of the first is [tex](A^T)^{-1}[/tex]. It's pretty easy to show for powers as well.