Section 8.7
Week 8
Ali Mousavidehshikh
Department of Mathematics University of Toronto
Ali Mousavidehshikh
Week 8
Outline
1 Section 8.7
Section 8.7
Week 8
Ali Mousavidehshikh
Section 8.7
Let i2 = −1. A complex number is a number of the form z = a + ib.
Week 8
Ali Mousavidehshikh
Section 8.7
Let i2 = −1. A complex number is a number of the form z = a + ib.
The conjugate of a complex number is z = a − bi.
Week 8
Ali Mousavidehshikh
Section 8.7
Let i2 = −1. A complex number is a number of the form z = a + ib.
The conjugate of a complex number is z = a − bi. We denote the set of all complex numbers by C.
Week 8
Ali Mousavidehshikh
Section 8.7
Let i2 = −1. A complex number is a number of the form z = a + ib.
The conjugate of a complex number is z = a − bi.
We denote the set of all complex numbers by C.
All the theorems proved so far for real vector spaces remain true for complex vector spaces (this is when the scalars are complex numbers instead of real numbers).
Ali Mousavidehshikh
Week 8
Section 8.7
Standard inner product in Cn: Let
z = (z1,z2,…,zn),w = (w1,w2,…,wn), where zi,wi ∈ C. We define the standard inner product ⟨z,w⟩ = ni=1 ziwi. Notice that if z , w ∈ Rn , then ⟨z , w ⟩ = z · w .
Week 8
Ali Mousavidehshikh
Section 8.7
Standard inner product in Cn: Let
z = (z1,z2,…,zn),w = (w1,w2,…,wn), where zi,wi ∈ C. We define the standard inner product ⟨z,w⟩ = ni=1 ziwi. Notice that if z , w ∈ Rn , then ⟨z , w ⟩ = z · w .
Let z = (3, 1 − 2i , 3 + i , 2i ) and w = (1 − i , −1, −i + 1, i ). Compute zw, z/w, ⟨z, w⟩, ⟨w, z⟩, ⟨z, z⟩, and ⟨w, w⟩.
Week 8
Ali Mousavidehshikh
Section 8.7
Standard inner product in Cn: Let
z = (z1,z2,…,zn),w = (w1,w2,…,wn), where zi,wi ∈ C. We define the standard inner product ⟨z,w⟩ = ni=1 ziwi. Notice that if z , w ∈ Rn , then ⟨z , w ⟩ = z · w .
Let z = (3, 1 − 2i , 3 + i , 2i ) and w = (1 − i , −1, −i + 1, i ). Compute zw, z/w, ⟨z, w⟩, ⟨w, z⟩, ⟨z, z⟩, and ⟨w, w⟩.
Theorem: Let z,z1,w,w1 ∈ Cn, and λ ∈ C. Then, (1) ⟨z + z1 , w ⟩ = ⟨z , w ⟩ + ⟨z1 + w ⟩, and
⟨z , w + w1 ⟩ = ⟨z , w ⟩ + ⟨z , w1 ⟩,
Ali Mousavidehshikh
Week 8
Section 8.7
Standard inner product in Cn: Let
z = (z1,z2,…,zn),w = (w1,w2,…,wn), where zi,wi ∈ C. We define the standard inner product ⟨z,w⟩ = ni=1 ziwi. Notice that if z , w ∈ Rn , then ⟨z , w ⟩ = z · w .
Let z = (3, 1 − 2i , 3 + i , 2i ) and w = (1 − i , −1, −i + 1, i ). Compute zw, z/w, ⟨z, w⟩, ⟨w, z⟩, ⟨z, z⟩, and ⟨w, w⟩.
Theorem: Let z,z1,w,w1 ∈ Cn, and λ ∈ C. Then, (1) ⟨z + z1 , w ⟩ = ⟨z , w ⟩ + ⟨z1 + w ⟩, and
⟨z , w + w1 ⟩ = ⟨z , w ⟩ + ⟨z , w1 ⟩,
(2) ⟨λz,w⟩ = λ⟨z,w⟩ and ⟨z,λw⟩ = λ⟨z,w⟩,
Ali Mousavidehshikh
Week 8
Section 8.7
Standard inner product in Cn: Let
z = (z1,z2,…,zn),w = (w1,w2,…,wn), where zi,wi ∈ C. We define the standard inner product ⟨z,w⟩ = ni=1 ziwi. Notice that if z , w ∈ Rn , then ⟨z , w ⟩ = z · w .
Let z = (3, 1 − 2i , 3 + i , 2i ) and w = (1 − i , −1, −i + 1, i ). Compute zw, z/w, ⟨z, w⟩, ⟨w, z⟩, ⟨z, z⟩, and ⟨w, w⟩.
Theorem: Let z,z1,w,w1 ∈ Cn, and λ ∈ C. Then, (1) ⟨z + z1 , w ⟩ = ⟨z , w ⟩ + ⟨z1 + w ⟩, and
⟨z , w + w1 ⟩ = ⟨z , w ⟩ + ⟨z , w1 ⟩,
(2) ⟨λz,w⟩ = λ⟨z,w⟩ and ⟨z,λw⟩ = λ⟨z,w⟩,
(3) ⟨z,w⟩ = ⟨w,z⟩,
Ali Mousavidehshikh
Week 8
Section 8.7
Standard inner product in Cn: Let
z = (z1,z2,…,zn),w = (w1,w2,…,wn), where zi,wi ∈ C. We define the standard inner product ⟨z,w⟩ = ni=1 ziwi. Notice that if z , w ∈ Rn , then ⟨z , w ⟩ = z · w .
Let z = (3, 1 − 2i , 3 + i , 2i ) and w = (1 − i , −1, −i + 1, i ). Compute zw, z/w, ⟨z, w⟩, ⟨w, z⟩, ⟨z, z⟩, and ⟨w, w⟩.
Theorem: Let z,z1,w,w1 ∈ Cn, and λ ∈ C. Then, (1) ⟨z + z1 , w ⟩ = ⟨z , w ⟩ + ⟨z1 + w ⟩, and
⟨z , w + w1 ⟩ = ⟨z , w ⟩ + ⟨z , w1 ⟩,
(2) ⟨λz,w⟩ = λ⟨z,w⟩ and ⟨z,λw⟩ = λ⟨z,w⟩,
(3) ⟨z,w⟩ = ⟨w,z⟩,
(4) ⟨z,z⟩ ≥ 0 and ⟨z,z⟩ = 0 if and only if z = 0.
Ali Mousavidehshikh
Week 8
Section 8.7
Norm or length of z is defined to be ∥ z ∥= ⟨z,z⟩.
Week 8
Ali Mousavidehshikh
Section 8.7
Norm or length of z is defined to be ∥ z ∥= ⟨z,z⟩. Find the norm of z = (1, 1 − i , 2 + 3i , −3i ).
Week 8
Ali Mousavidehshikh
Section 8.7
Norm or length of z is defined to be ∥ z ∥= ⟨z,z⟩. Find the norm of z = (1, 1 − i , 2 + 3i , −3i ).
Theorem: For any z ∈ Cn,
(1) ∥ z ∥≥ 0, and ∥ z ∥= 0 if and only if z = 0,
Week 8
Ali Mousavidehshikh
Section 8.7
Norm or length of z is defined to be ∥ z ∥= ⟨z,z⟩. Find the norm of z = (1, 1 − i , 2 + 3i , −3i ).
Theorem: For any z ∈ Cn,
(1) ∥ z ∥≥ 0, and ∥ z ∥= 0 if and only if z = 0,
(2) ∥ λz ∥= |λ| ∥ z ∥ for all complex numbers λ (where |λ| is the norm of λ).
Week 8
Ali Mousavidehshikh
Section 8.7
Norm or length of z is defined to be ∥ z ∥= ⟨z,z⟩. Find the norm of z = (1, 1 − i , 2 + 3i , −3i ).
Theorem: For any z ∈ Cn,
(1) ∥ z ∥≥ 0, and ∥ z ∥= 0 if and only if z = 0,
(2) ∥ λz ∥= |λ| ∥ z ∥ for all complex numbers λ (where |λ| is the norm of λ).
Avectoru∈Cn iscalledaunitvectorif∥u∥=1. Thevector
u= z isaunitvectorforanynon-zeroz∈Cn. ∥z∥
Week 8
Ali Mousavidehshikh
Section 8.7
Norm or length of z is defined to be ∥ z ∥= ⟨z,z⟩. Find the norm of z = (1, 1 − i , 2 + 3i , −3i ).
Theorem: For any z ∈ Cn,
(1) ∥ z ∥≥ 0, and ∥ z ∥= 0 if and only if z = 0,
(2) ∥ λz ∥= |λ| ∥ z ∥ for all complex numbers λ (where |λ| is the norm of λ).
Avectoru∈Cn iscalledaunitvectorif∥u∥=1. Thevector u= z isaunitvectorforanynon-zeroz∈Cn.
∥z∥
A matrix A = [aij ] is called a complex matrix if aij is a complex number for all i,j. We define its conjugate to be A = [aij ]. Notice that A + B = A + B and AB = AB .
Ali Mousavidehshikh
Week 8
Section 8.7
The transpose conjugate of A is denoted by AH, that is, AH =(A)T =AT.
Week 8
Ali Mousavidehshikh
Section 8.7
The transpose conjugate of A is denoted by AH, that is, AH =(A)T =AT.
A = A if and only if A is a real matrix.
Week 8
Ali Mousavidehshikh
Section 8.7
The transpose conjugate of A is denoted by AH, that is, AH =(A)T =AT.
A = A if and only if A is a real matrix.
Observe that AH = AT if and only if A is a real matrix.
Week 8
Ali Mousavidehshikh
Section 8.7
The transpose conjugate of A is denoted by AH, that is, AH =(A)T =AT.
A = A if and only if A is a real matrix.
Observe that AH = AT if and only if A is a real matrix.
3 2i−4+i1−i Example: LetA= −3+i 1+i −4 2i . FindA
and AH.
Week 8
Ali Mousavidehshikh
Section 8.7
The transpose conjugate of A is denoted by AH, that is, AH =(A)T =AT.
A = A if and only if A is a real matrix.
Observe that AH = AT if and only if A is a real matrix.
3 2i−4+i1−i Example: LetA= −3+i 1+i −4 2i . FindA
and AH.
Theorem: (1) (AH )H = A,
Week 8
Ali Mousavidehshikh
Section 8.7
The transpose conjugate of A is denoted by AH, that is, AH =(A)T =AT.
A = A if and only if A is a real matrix.
Observe that AH = AT if and only if A is a real matrix.
3 2i−4+i1−i Example: LetA= −3+i 1+i −4 2i . FindA
and AH.
Theorem: (1) (AH )H = A, (2)(A+B)H =AH +BH,
Week 8
Ali Mousavidehshikh
Section 8.7
The transpose conjugate of A is denoted by AH, that is, AH =(A)T =AT.
A = A if and only if A is a real matrix.
Observe that AH = AT if and only if A is a real matrix.
3 2i−4+i1−i Example: LetA= −3+i 1+i −4 2i . FindA
and AH.
Theorem: (1) (AH )H = A, (2)(A+B)H =AH +BH, (3) (λA)H = λAH ,
Week 8
Ali Mousavidehshikh
Section 8.7
The transpose conjugate of A is denoted by AH, that is, AH =(A)T =AT.
A = A if and only if A is a real matrix.
Observe that AH = AT if and only if A is a real matrix.
3 2i−4+i1−i Example: LetA= −3+i 1+i −4 2i . FindA
and AH.
Theorem: (1) (AH )H = A, (2)(A+B)H =AH +BH, (3) (λA)H = λAH , (4)(AB)H =BHAH.
Week 8
Ali Mousavidehshikh
Section 8.7
The transpose conjugate of A is denoted by AH, that is, AH =(A)T =AT.
A = A if and only if A is a real matrix.
Observe that AH = AT if and only if A is a real matrix.
3 2i−4+i1−i Example: LetA= −3+i 1+i −4 2i . FindA
and AH.
Theorem: (1) (AH )H = A, (2)(A+B)H =AH +BH, (3) (λA)H = λAH , (4)(AB)H =BHAH.
Definition: A square matrix A is called hermitian if AH = A, equivalently A = AT .
Ali Mousavidehshikh
Week 8
hermitian matrix.
Section 8.7
Hermitian matrices are easy to recognize because the entries on the main diagonal must be real, and the reflection of each non-diagonal entry in the main diagonal must be the
3 −i2+i conjugate of that entry: A = i 1 −6 is a
2−i −6 −4
Week 8
Ali Mousavidehshikh
hermitian matrix.
⟨Az,w⟩ = ⟨z,Aw⟩ for all z,w ∈ Cn.
Section 8.7
Hermitian matrices are easy to recognize because the entries on the main diagonal must be real, and the reflection of each non-diagonal entry in the main diagonal must be the
3 −i2+i conjugate of that entry: A = i 1 −6 is a
2−i −6 −4
An n × n complex matrix A is hermitian if and only if
Week 8
Ali Mousavidehshikh
Section 8.7
Hermitian matrices are easy to recognize because the entries on the main diagonal must be real, and the reflection of each non-diagonal entry in the main diagonal must be the
3 −i2+i conjugate of that entry: A = i 1 −6 is a
2−i −6 −4
An n × n complex matrix A is hermitian if and only if
hermitian matrix.
⟨Az,w⟩ = ⟨z,Aw⟩ for all z,w ∈ Cn.
If A is an n × n complex matrix, then λ is an eigenvalue of A if Ax = λx holds for some non-zero column x ∈ C. In this case x is called an eigenvector corresponding to λ. The characteristic polynomial cA(x) is defined by
cA(x) = det(xI − A). This polynomial has complex coefficients. All the proofs from the real case carry over.
Week 8
Ali Mousavidehshikh
Section 8.7
Hermitian matrices are easy to recognize because the entries on the main diagonal must be real, and the reflection of each non-diagonal entry in the main diagonal must be the
3 −i2+i conjugate of that entry: A = i 1 −6 is a
2−i −6 −4
An n × n complex matrix A is hermitian if and only if
hermitian matrix.
⟨Az,w⟩ = ⟨z,Aw⟩ for all z,w ∈ Cn.
If A is an n × n complex matrix, then λ is an eigenvalue of A if Ax = λx holds for some non-zero column x ∈ C. In this case x is called an eigenvector corresponding to λ. The characteristic polynomial cA(x) is defined by
cA(x) = det(xI − A). This polynomial has complex coefficients. All the proofs from the real case carry over.
We say that two vectors z , w ∈ Cn are orthogonal if ⟨z,w⟩ = 0.
Ali Mousavidehshikh
Week 8
Section 8.7
Theorem: Let A denote a hermitian matrix. (1) The eigenvalues of A are real,
Week 8
Ali Mousavidehshikh
Section 8.7
Theorem: Let A denote a hermitian matrix.
(1) The eigenvalues of A are real,
(2) Eigenvectors of A corresponding to distinct eigenvalues are orthogonal (in general they are linearly independent).
Week 8
Ali Mousavidehshikh
Section 8.7
Theorem: Let A denote a hermitian matrix.
(1) The eigenvalues of A are real,
(2) Eigenvectors of A corresponding to distinct eigenvalues are orthogonal (in general they are linearly independent).
The definition of orthogonal sets and orthonormal sets are the same as in Rn.
Week 8
Ali Mousavidehshikh
Section 8.7
Theorem: Let A denote a hermitian matrix.
(1) The eigenvalues of A are real,
(2) Eigenvectors of A corresponding to distinct eigenvalues are orthogonal (in general they are linearly independent).
The definition of orthogonal sets and orthonormal sets are the same as in Rn.
Theorem: The following are equivalent for an n × n square matrix A.
(1) A is invertible and A−1 = AT .
Week 8
Ali Mousavidehshikh
Section 8.7
Theorem: Let A denote a hermitian matrix.
(1) The eigenvalues of A are real,
(2) Eigenvectors of A corresponding to distinct eigenvalues are orthogonal (in general they are linearly independent).
The definition of orthogonal sets and orthonormal sets are the same as in Rn.
Theorem: The following are equivalent for an n × n square matrix A.
(1) A is invertible and A−1 = AT .
(2) The rows of A are an orthonormal set in Cn.
Week 8
Ali Mousavidehshikh
Section 8.7
Theorem: Let A denote a hermitian matrix.
(1) The eigenvalues of A are real,
(2) Eigenvectors of A corresponding to distinct eigenvalues are orthogonal (in general they are linearly independent).
The definition of orthogonal sets and orthonormal sets are the same as in Rn.
Theorem: The following are equivalent for an n × n square matrix A.
(1) A is invertible and A−1 = AT .
(2) The rows of A are an orthonormal set in Cn.
(3) The columns of A are an orthonormal set in Cn.
Week 8
Ali Mousavidehshikh
Section 8.7
Theorem: Let A denote a hermitian matrix.
(1) The eigenvalues of A are real,
(2) Eigenvectors of A corresponding to distinct eigenvalues are orthogonal (in general they are linearly independent).
The definition of orthogonal sets and orthonormal sets are the same as in Rn.
Theorem: The following are equivalent for an n × n square matrix A.
(1) A is invertible and A−1 = AT .
(2) The rows of A are an orthonormal set in Cn.
(3) The columns of A are an orthonormal set in Cn. Definition. A square matrix is called unitary if U−1 = UH.
Ali Mousavidehshikh
Week 8
Section 8.7
Schur’s Theorem: If A is any n × n complex matrix, there exists a unitary matrix U such that UHAU = T, where T is upper triangular. Moreover, the entries on the main diagonal of T are the eigenvalues of A (including multiplicities).
Week 8
Ali Mousavidehshikh
Section 8.7
Schur’s Theorem: If A is any n × n complex matrix, there exists a unitary matrix U such that UHAU = T, where T is upper triangular. Moreover, the entries on the main diagonal of T are the eigenvalues of A (including multiplicities).
An n × n complex matrix A is called unitary diagonalizable if UHAU is diagonal for some unitary matrix U.
Week 8
Ali Mousavidehshikh
Section 8.7
Schur’s Theorem: If A is any n × n complex matrix, there exists a unitary matrix U such that UHAU = T, where T is upper triangular. Moreover, the entries on the main diagonal of T are the eigenvalues of A (including multiplicities).
An n × n complex matrix A is called unitary diagonalizable if UHAU is diagonal for some unitary matrix U.
Let A be an n×n complex matrix, and let λ1,λ2,…,λn denote the eigenvalues of A, including multiplicity. Then detA = λ1λ2 ···λn and tr A = ni=1 λi.
Week 8
Ali Mousavidehshikh
Section 8.7
Schur’s Theorem: If A is any n × n complex matrix, there exists a unitary matrix U such that UHAU = T, where T is upper triangular. Moreover, the entries on the main diagonal of T are the eigenvalues of A (including multiplicities).
An n × n complex matrix A is called unitary diagonalizable if UHAU is diagonal for some unitary matrix U.
Let A be an n×n complex matrix, and let λ1,λ2,…,λn denote the eigenvalues of A, including multiplicity. Then detA = λ1λ2 ···λn and tr A = ni=1 λi.
Schur’s Theorem tells us that every n × n complex matrix can 1 1
be unitarily triangularized. Notice that A = 0 1 cannot
be unitarily diagonalized. That is, there does not exist a unitary matrix U such that U−1AU is a diagonal matrix (to the students: try doing this).
Ali Mousavidehshikh
Week 8
Section 8.7
Spectral Theorem: If A is hermitian, then A is unitarily diagonalizable.
Week 8
Ali Mousavidehshikh
Section 8.7
Spectral Theorem: If A is hermitian, then A is unitarily diagonalizable.
3 2+i
Example: LetA= 2−i 7 . NoticethatU is
hermitian. Find a unitary matrix U such that UHAU is a diagonal matrix.
Week 8
Ali Mousavidehshikh
Section 8.7
Spectral Theorem: If A is hermitian, then A is unitarily diagonalizable.
3 2+i
Example: LetA= 2−i 7 . NoticethatU is
hermitian. Find a unitary matrix U such that UHAU is a diagonal matrix.
Solution. The characteristic polynomial of A is
CA(x) = det(xI − A) = (x − 2)(x − 8) (both real as expected, as A is hermitian).
Week 8
Ali Mousavidehshikh
Section 8.7
Spectral Theorem: If A is hermitian, then A is unitarily diagonalizable.
3 2+i
Example: LetA= 2−i 7 . NoticethatU is
hermitian. Find a unitary matrix U such that UHAU is a diagonal matrix.
Solution. The characteristic polynomial of A is
CA(x) = det(xI − A) = (x − 2)(x − 8) (both real as expected, as A is hermitian). The eigenvectors corresponding to 2 and 8
2+i 1
are −2 and 2 − i (orthogonal as expected),
respectively.
Week 8
Ali Mousavidehshikh
Section 8.7
Spectral Theorem: If A is hermitian, then A is unitarily diagonalizable.
3 2+i
Example: LetA= 2−i 7 . NoticethatU is
hermitian. Find a unitary matrix U such that UHAU is a diagonal matrix.
Solution. The characteristic polynomial of A is
CA(x) = det(xI − A) = (x − 2)(x − 8) (both real as expected, as A is hermitian). The eigenvectors corresponding to 2 and 8
2+i 1
are −2 and 2 − i (orthogonal as expected),
√
respectively. Each of these eigenvectors has norm
6.
Week 8
Ali Mousavidehshikh
Section 8.7
Spectral Theorem: If A is hermitian, then A is unitarily diagonalizable.
3 2+i
Example: LetA= 2−i 7 . NoticethatU is
hermitian. Find a unitary matrix U such that UHAU is a diagonal matrix.
Solution. The characteristic polynomial of A is
CA(x) = det(xI − A) = (x − 2)(x − 8) (both real as expected, as A is hermitian). The eigenvectors corresponding to 2 and 8
2+i 1
are −2 and 2 − i (orthogonal as expected),
√
respectively. Each of these eigenvectors has norm 12+i 1
6. Letting
U=√6 −12−i.
Week 8
Ali Mousavidehshikh
Section 8.7
Spectral Theorem: If A is hermitian, then A is unitarily diagonalizable.
3 2+i
Example: LetA= 2−i 7 . NoticethatU is
hermitian. Find a unitary matrix U such that UHAU is a diagonal matrix.
Solution. The characteristic polynomial of A is
CA(x) = det(xI − A) = (x − 2)(x − 8) (both real as expected, as A is hermitian). The eigenvectors corresponding to 2 and 8
2+i 1
are −2 and 2 − i (orthogonal as expected),
√
U = √6 −1 2 − i . Then U is unitary (rows/columns are
respectively. Each of these eigenvectors has norm 12+i 1
6. Letting
H 20 orthonormal) and U AU = 0 8 .
Week 8
Ali Mousavidehshikh
Section 8.7
Spectral Theorem: If A is hermitian, then A is unitarily diagonalizable.
3 2+i
Example: LetA= 2−i 7 . NoticethatU is
hermitian. Find a unitary matrix U such that UHAU is a diagonal matrix.
Solution. The characteristic polynomial of A is
CA(x) = det(xI − A) = (x − 2)(x − 8) (both real as expected, as A is hermitian). The eigenvectors corresponding to 2 and 8
2+i 1
are −2 and 2 − i (orthogonal as expected),
√
U = √6 −1 2 − i . Then U is unitary (rows/columns are
respectively. Each of these eigenvectors has norm 12+i 1
6. Letting
H 20 orthonormal) and U AU = 0 8 .
In the real case, we know a matrix is orthogonally diagonalizable if and only if it is symmetric. The spectral
theorem is the analog of half
Ali Mousavidehshikh
of this result (hermitian implies
Week 8
Section 8.7
In the real case, we know a matrix is orthogonally diagonalizable if and only if it is symmetric. The spectral theorem is the analog of half of this result (hermitian implies unitarily diagonalizable). However, unlike the real case, the
0 1 converse fails for complex matrices. The matrix −1 0
non-hermitian. However, it is unitarily diagonalizable.
is
Week 8
Ali Mousavidehshikh
Section 8.7
In the real case, we know a matrix is orthogonally diagonalizable if and only if it is symmetric. The spectral theorem is the analog of half of this result (hermitian implies unitarily diagonalizable). However, unlike the real case, the
0 1 converse fails for complex matrices. The matrix −1 0 is
non-hermitian. However, it is unitarily diagonalizable.
Definition: A square complex matrix N is called normal if NNH =NHN.
Week 8
Ali Mousavidehshikh
Section 8.7
In the real case, we know a matrix is orthogonally diagonalizable if and only if it is symmetric. The spectral theorem is the analog of half of this result (hermitian implies unitarily diagonalizable). However, unlike the real case, the
0 1 converse fails for complex matrices. The matrix −1 0 is
non-hermitian. However, it is unitarily diagonalizable. Definition: A square complex matrix N is called normal if
NNH =NHN.
Theorem: A square complex matrix N is diagonalizable if and only if N is normal.
Week 8
Ali Mousavidehshikh
Section 8.7
In the real case, we know a matrix is orthogonally diagonalizable if and only if it is symmetric. The spectral theorem is the analog of half of this result (hermitian implies unitarily diagonalizable). However, unlike the real case, the
0 1 converse fails for complex matrices. The matrix −1 0 is
non-hermitian. However, it is unitarily diagonalizable. Definition: A square complex matrix N is called normal if
NNH =NHN.
Theorem: A square complex matrix N is diagonalizable if and
only if N is normal.
Cayley-Hamilton Theorem: If A is an n × n complex matrix, then CA(A) = 0; that is, A is a root of its characteristic polynomial.
Ali Mousavidehshikh
Week 8