Let's say we have the square nn x nn matrix AA
composed of real numbers. By "square", we mean it has the same number
of rows as columns.
A=(
a1,1…a1,n
⋮⋱⋮
an,1…an,n
)
A
a11
…
a1n
⋮
⋱
⋮
an1
…
ann
(1)
The subscripts of the real numbers in this matrix denote the row and
column numbers, respectively (i.e.
a1,2a12
holds the position at the intersection of the first row and the
second column).
We will denote the inverse of this matrix as
A-1
A
.
A matrix inverse has the property that when it is multiplied by the
original matrix (on the left or on the right), the result will be the
identity matrix.
AA-1=A-1A=I
A
A
A
A
I
(2)
To compute the inverse of AA, two steps are required. Both involve
taking determinants (
detA
A
)
of matrices. The first step is to find the adjoint (
AH
A
)
of the matrix A. It is computed as follows:
AH=(
α1,1…α1,n
⋮⋱⋮
αn,1…αn,n
)
A
α11
…
α1n
⋮
⋱
⋮
αn1
…
αnn
(3)
αi,j=-1i+jdet
A
ij
αij
-1
i
j
A
ij
(4)
where
A
ij
A
ij
is the (n−1)n1 x (n−1)n1 matrix obtained from AA by eliminating its ii-th column and jj-th row. Note that we are not eliminating the ii-th row and jj-th column as you might expect.
To finish the process of determining the inverse, simply divide the
adjoint by the determinant of the original matrix AA .
A=(
ab
cd
)
A
a
b
c
d
(6)
Find the inverse of the above matrix.
The first step is to compute the terms in the adjoint matrix:
α1,1=-12d
α11
-1
2
d
(7)
α1,2=-13b
α12
-1
3
b
(8)
α2,1=-13c
α21
-1
3
c
(9)
α2,2=-14a
α22
-1
4
a
(10)
Therefore,
AH=(
d−b
−ca
)
A
d
b
c
a
(11)
We then compute the determinant to be
ad−bc
a
d
b
c
.
Dividing through by this quantity yields the inverse of AA:
A-1=1ad−bc(
d−b
−ca
)
A
1
a
d
b
c
d
b
c
a
(12)