Section 4.2

Supplementary Problems

[43/0%]
LaTeX
Problem 4.2.1

Let AA, BB be 2×22 \times 2 matrices over C\mathbb {C}. What are the conditions on AA, BB such that tr⁡(AB)=tr⁡(A)tr⁡(B)\operatorname {tr}(AB) = \operatorname {tr}(A)\operatorname {tr}(B)?

?
Problem 4.2.2

Given a 2×22 \times 2 matrix MM over C\mathbb {C}. What can be said about the determinant of M⊗I2−I2⊗MM \otimes I_{2} - I_{2} \otimes M?

?
Problem 4.2.3
?
(i)

What is condition on α∈R\alpha \in \mathbb {R} such that

F(α)=(1ααα1ααα1) F(\alpha ) = \begin{pmatrix} 1 & \alpha & \alpha \\ \alpha & 1 & \alpha \\ \alpha & \alpha & 1 \end{pmatrix}

is invertible?

(ii)

What is the condition on α,β∈R\alpha , \beta \in \mathbb {R} such that

G(α,β)=(1ββββ1ααβα1αβαα1) G(\alpha ,\beta ) = \begin{pmatrix} 1 & \beta & \beta & \beta \\ \beta & 1 & \alpha & \alpha \\ \beta & \alpha & 1 & \alpha \\ \beta & \alpha & \alpha & 1 \end{pmatrix}

is invertible?

Problem 4.2.4

Let AA, BB be n×nn \times n matrices over C\mathbb {C}. Then tr⁡([A,B])=0\operatorname {tr}([A,B]) = 0. Consider n=3n=3 and

C=(10000000−1) C = \begin{pmatrix} 1 & 0 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & -1 \end{pmatrix}

with tr⁡(C)=0\operatorname {tr}(C) = 0. Construct 3×33 \times 3 matrices AA and BB such that [A,B]=C[A,B] = C.

?
Problem 4.2.5

Consider the kk vectors

v→j=(v1jv2j⋮vnj)j=1,…,k \overrightarrow {v}_{j} = \begin{pmatrix} v_{1j} \\ v_{2j} \\ \vdots \\ v_{nj} \end{pmatrix} \qquad j = 1, \ldots , k

in the Euclidean space Rn\mathbb {R}^{n}. The parallelepiped determined by these vectors has the kk-dimensional area

det⁡(VTV) \sqrt{\operatorname {det}\left(V^{T}V\right)}

where VV is the n×kn \times k matrix with v→1,v→2,…,v→k\overrightarrow {v}_{1}, \overrightarrow {v}_{2}, \ldots , \overrightarrow {v}_{k} as its columns. Apply it to the two vectors in R3\mathbb {R}^{3}

v→1=12(101),v→2=12(10−1). \overrightarrow {v}_{1} = \frac{1}{\sqrt{2}} \begin{pmatrix} 1 \\ 0 \\ 1 \end{pmatrix}, \qquad \overrightarrow {v}_{2} = \frac{1}{\sqrt{2}} \begin{pmatrix} 1 \\ 0 \\ -1 \end{pmatrix}.
?
Problem 4.2.6

The equation of a hyperplane passing through the nn points x→1,x→2,…,x→n\overrightarrow {x}_{1}, \overrightarrow {x}_{2}, \ldots , \overrightarrow {x}_{n} in the Euclidean space Rn\mathbb {R}^{n} is given by

det⁡(111…1x→x→1x→2…x→n)=0. \operatorname {det}\left(\begin{array}{lllll} 1 & 1 & 1 & \ldots & 1 \\ \overrightarrow {x} & \overrightarrow {x}_{1} & \overrightarrow {x}_{2} & \ldots & \overrightarrow {x}_{n} \end{array}\right) = 0.

Let n=3n=3. Apply it to the three vectors in R3\mathbb {R}^{3}

x→1=12(101),x→2=(010),x→3=12(101). \overrightarrow {x}_{1} = \frac{1}{\sqrt{2}} \begin{pmatrix} 1 \\ 0 \\ 1 \end{pmatrix}, \qquad \overrightarrow {x}_{2} = \begin{pmatrix} 0 \\ 1 \\ 0 \end{pmatrix}, \qquad \overrightarrow {x}_{3} = \frac{1}{\sqrt{2}} \begin{pmatrix} 1 \\ 0 \\ 1 \end{pmatrix}.
?
Problem 4.2.7

Let AA, BB be 2×22 \times 2 matrices. Show that

[A,B]+≡AB+BA≡(tr⁡(AB)−tr⁡(A)tr⁡(B))I2+tr⁡(A)B+tr⁡(B)A. [A,B]_{+} \equiv AB+BA \equiv (\operatorname {tr}(AB)-\operatorname {tr}(A)\operatorname {tr}(B))I_{2} + \operatorname {tr}(A)B + \operatorname {tr}(B)A.

Can this identity be extended to 3×33 \times 3 matrices?

?
Problem 4.2.8
?
(i)

Find all nonzero 2×22 \times 2 matrices AA and BB such that BA∗=tr⁡(BA∗)ABA^{*} = \operatorname {tr}(BA^{*})A.

(ii)

Find all 2×22 \times 2 matrices AA such that A2=tr⁡(A)AA^{2} = \operatorname {tr}(A)A. Calculate det⁡(A)\operatorname {det}\left(A\right) and det⁡(A2)\operatorname {det}\left(A^{2}\right) of such a matrix.

(iii)

Let BB be a 2×22 \times 2 matrix over C\mathbb {C}. Assume that B2=I2B^{2} = I_{2} and thus tr⁡(B2)=2\operatorname {tr}(B^{2}) = 2. What can be said about the trace of BB?

(iv)

Find all 2×22 \times 2 matrices CC with tr⁡(C2)−(tr⁡(C))2=0\operatorname {tr}(C^{2})-(\operatorname {tr}(C))^{2} = 0.

(v)

Let DD be a 2×22 \times 2 matrix over C\mathbb {C}. Find the condition on DD such that det⁡(D)=tr⁡(D2)\operatorname {det}\left(D\right) = \operatorname {tr}(D^{2}). Find the condition on DD such that det⁡(D⊗D)=tr⁡(D⊗D)\operatorname {det}\left(D \otimes D\right) = \operatorname {tr}(D \otimes D).

Problem 4.2.9

Let AA, BB be n×nn \times n matrices. Let UU, VV be unitary n×nn \times n matrices. Show that

tr⁡((U⊗V)(A⊗B)(U∗⊗V∗))=tr⁡(A)tr⁡(B) \operatorname {tr}((U \otimes V)(A \otimes B)(U^{*} \otimes V^{*})) = \operatorname {tr}(A)\operatorname {tr}(B)

applying cyclic invariance and UU∗=InUU^{*} = I_{n}, VV∗=InVV^{*} = I_{n}.

?
Problem 4.2.10

Let AA, BB be 2×22 \times 2 matrices over C\mathbb {C} with tr⁡(A)=0\operatorname {tr}(A) = 0 and tr⁡(B)=0\operatorname {tr}(B) = 0. Show that the condition on AA and BB such that tr⁡(AB)=0\operatorname {tr}(AB) = 0 is given by

2a11b11+a12b21+a21b12=0. 2a_{11}b_{11} + a_{12}b_{21} + a_{21}b_{12} = 0.
?
Problem 4.2.11

Let AA, BB be nonzero 2×22 \times 2 matrices over R\mathbb {R} with det⁡(A⊗B)=0\operatorname {det}\left(A \otimes B\right) = 0. Can we conclude that both AA and BB are noninvertible?

?
Problem 4.2.12

Consider the symmetric n×nn \times n band matrix (n≥3n \geq 3)

Mn=(110…00111…00011…00⋮⋮⋮⋱⋮⋮000…11000…11) M_{n} = \begin{pmatrix} 1 & 1 & 0 & \ldots & 0 & 0 \\ 1 & 1 & 1 & \ldots & 0 & 0 \\ 0 & 1 & 1 & \ldots & 0 & 0 \\ \vdots & \vdots & \vdots & \ddots & \vdots & \vdots \\ 0 & 0 & 0 & \ldots & 1 & 1 \\ 0 & 0 & 0 & \ldots & 1 & 1 \end{pmatrix}

with the elements in the field F2\mathbb {F}_{2}. Show that

det⁡(Mn)=det⁡(Mn−1)−det⁡(Mn−2) \operatorname {det}\left(M_{n}\right) = \operatorname {det}\left(M_{n-1}\right) - \operatorname {det}\left(M_{n-2}\right)

with the initial conditions det⁡(M3)=det⁡(M4)=1\operatorname {det}\left(M_{3}\right) = \operatorname {det}\left(M_{4}\right) = 1. Show that the solution is

det⁡(Mn)=233cos⁡(nπ3−π6)(mod2). \operatorname {det}\left(M_{n}\right) = \frac{2\sqrt{3}}{3} \cos \left(\frac{n\pi }{3} - \frac{\pi }{6}\right) \pmod{2}.
?
Problem 4.2.13

Consider the nonnormal matrices

A2=(0100),A3=(011001000). A_{2} = \begin{pmatrix} 0 & 1 \\ 0 & 0 \end{pmatrix}, \qquad A_{3} = \begin{pmatrix} 0 & 1 & 1 \\ 0 & 0 & 1 \\ 0 & 0 & 0 \end{pmatrix}.

Find (det⁡(A2∗A2))1/2(\operatorname {det}\left(A_{2}^{*}A_{2}\right))^{1/2} and (det⁡(A3∗A3))1/2(\operatorname {det}\left(A_{3}^{*}A_{3}\right))^{1/2}. Extend to nn dimensions.

?
Problem 4.2.14

Let AA be a n×nn \times n matrix with det⁡(A)=−1\operatorname {det}\left(A\right) = -1. Show that det⁡(A−1)=−1\operatorname {det}\left(A^{-1}\right) = -1.

?
Problem 4.2.15

Let nn be even. Consider a skew-symmetric n×nn \times n matrix AA over R\mathbb {R}. Let BB be a symmetric n×nn \times n matrix over R\mathbb {R} with the entries of B=(bjk)B = (b_{jk}) given by bjk=bjbkb_{jk} = b_{j}b_{k} (bj,bk∈Rb_{j}, b_{k} \in \mathbb {R}). Show that

det⁡(A+B)=det⁡(A). \operatorname {det}\left(A+B\right) = \operatorname {det}\left(A\right).

Notice that det⁡(B)=0\operatorname {det}\left(B\right) = 0. Let SnS_{n} be the symmetric group. Then we have

det⁡(A+B)=∑σ∈Sn(−1)σ(a1σ(1)+b1bσ(1))(a2σ(2)+b2bσ(2))⋯(anσ(n)+bnbσ(n)). \operatorname {det}\left(A+B\right) = \sum _{\sigma \in S_{n}} (-1)^{\sigma } (a_{1\sigma (1)}+b_{1}b_{\sigma (1)})(a_{2\sigma (2)}+b_{2}b_{\sigma (2)}) \cdots (a_{n\sigma (n)}+b_{n}b_{\sigma (n)}).
?
Problem 4.2.16

Let h>0h>0 and b>ab>a. Consider the matrix

M(a,b,h)=(1ah1bh). M(a,b,h) = \begin{pmatrix} 1 & ah \\ 1 & bh \end{pmatrix}.

Show that det⁡(M)=h(b−a)>0\operatorname {det}\left(M\right) = h(b-a) > 0 and the inverse of MM is given by

M−1(a,b,h)=(b/(b−a)−a/(b−a)−1/((b−a)h)1/((b−a)h)). M^{-1}(a,b,h) = \begin{pmatrix} b/(b-a) & -a/(b-a) \\ -1/((b-a)h) & 1/((b-a)h) \end{pmatrix}.
?
Problem 4.2.17

The n×nn \times n permutation matrices form a group under matrix multiplications. Show that det⁡(In−P)=0\operatorname {det}\left(I_{n}-P\right) = 0 for any n×nn \times n permutation matrix.

?
Problem 4.2.18

Let u→,v→,w→\overrightarrow {u}, \overrightarrow {v}, \overrightarrow {w} be column vectors in R3\mathbb {R}^{3}. We form the 3×33 \times 3 matrix

M=(u→v→w→). M = \begin{pmatrix} \overrightarrow {u} & \overrightarrow {v} & \overrightarrow {w} \end{pmatrix}.

Show that det⁡(M)=(u→×v→)⋅w→\operatorname {det}\left(M\right) = (\overrightarrow {u} \times \overrightarrow {v}) \cdot \overrightarrow {w}.

?
Problem 4.2.19
?
(i)

Let n≥2n \geq 2. Consider the n×nn \times n matrix

A=(123…n−1n234…n1345…12⋮⋮⋮⋱⋮⋮n12…n−2n−1). A = \begin{pmatrix} 1 & 2 & 3 & \ldots & n-1 & n \\ 2 & 3 & 4 & \ldots & n & 1 \\ 3 & 4 & 5 & \ldots & 1 & 2 \\ \vdots & \vdots & \vdots & \ddots & \vdots & \vdots \\ n & 1 & 2 & \ldots & n-2 & n-1 \end{pmatrix}.

Show that det⁡(A)=(−1)n(n−1)/212(n+1)nn−1\operatorname {det}\left(A\right) = (-1)^{n(n-1)/2} \frac{1}{2}(n+1)n^{n-1}.

(ii)

Let n≥2n \geq 2. Consider the n×nn \times n matrix

M(x)=(c1xx…xxxc2x…xxxxc3…xx⋮⋮⋮⋱⋮⋮xxx…xcn). M(x) = \begin{pmatrix} c_{1} & x & x & \ldots & x & x \\ x & c_{2} & x & \ldots & x & x \\ x & x & c_{3} & \ldots & x & x \\ \vdots & \vdots & \vdots & \ddots & \vdots & \vdots \\ x & x & x & \ldots & x & c_{n} \end{pmatrix}.

Show that det⁡(M(x))=(−1)n(P(x)−xdP(x)/dx)\operatorname {det}\left(M(x)\right) = (-1)^{n}(P(x)-xdP(x)/dx), where

P(x)=(x−c1)(x−c2)⋯(x−cn). P(x) = (x-c_{1})(x-c_{2}) \cdots (x-c_{n}).
Problem 4.2.20

Consider the 2×22 \times 2 matrix

C=(0111)  ⇒  tr⁡(C)=1,  tr⁡(C2)=3. C = \begin{pmatrix} 0 & 1 \\ 1 & 1 \end{pmatrix} \; \Rightarrow \; \operatorname {tr}(C) = 1, \; \operatorname {tr}(C^{2}) = 3.

Show that tr⁡(Ck)=tr⁡(Ck−1)+tr⁡(Ck−2)\operatorname {tr}(C^{k}) = \operatorname {tr}(C^{k-1}) + \operatorname {tr}(C^{k-2}) with k=3,4,…k=3,4,\ldots

?
Problem 4.2.21

Let AA be an n×nn \times n matrix. Assume that

tr⁡(Aj)=0forj=1,…,n. \operatorname {tr}(A^{j}) = 0 \qquad \text{for} \quad j=1,\ldots ,n.

Can we conclude that det⁡(A)=0\operatorname {det}\left(A\right) = 0?

?
Problem 4.2.22

Let A1A_{1}, A2A_{2}, A3A_{3} be 2×22 \times 2 matrices over R\mathbb {R}. We define

cj=tr⁡(Aj),j=1,2,3;cjk=tr⁡(AjAk),j,k=1,2,3. c_{j} = \operatorname {tr}(A_{j}), \quad j=1,2,3; \qquad c_{jk} = \operatorname {tr}(A_{j}A_{k}), \quad j,k=1,2,3.
?
(i)

Given the coefficients cjc_{j}, cjkc_{jk} reconstruct the matrices A1A_{1}, A2A_{2}, A3A_{3}.

(ii)

Apply the result to the case c1=c2=c3=0c_{1}=c_{2}=c_{3}=0

c11=0,  c12=1,  c13=0,  c21=1,  c22=0,  c23=0,  c31=0,  c32=0,  c33=2 c_{11}=0, \; c_{12}=1, \; c_{13}=0, \; c_{21}=1, \; c_{22}=0, \; c_{23}=0, \; c_{31}=0, \; c_{32}=0, \; c_{33}=2

and show that

A1=(0100),A2=(0010),A3=(100−1). A_{1} = \begin{pmatrix} 0 & 1 \\ 0 & 0 \end{pmatrix}, \qquad A_{2} = \begin{pmatrix} 0 & 0 \\ 1 & 0 \end{pmatrix}, \qquad A_{3} = \begin{pmatrix} 1 & 0 \\ 0 & -1 \end{pmatrix}.
Problem 4.2.23
?
(i)

Let BB be a 2×22 \times 2 matrix over C\mathbb {C}. Given

d1=det⁡(B),d2=det⁡(B2),d3=det⁡(B3),d4=det⁡(B4). d_{1} = \operatorname {det}\left(B\right), \quad d_{2} = \operatorname {det}\left(B^{2}\right), \quad d_{3} = \operatorname {det}\left(B^{3}\right), \quad d_{4} = \operatorname {det}\left(B^{4}\right).

Can we reconstruct BB from d1,d2,d3,d4d_{1}, d_{2}, d_{3}, d_{4}. Does it depend on whether the matrix BB is normal?

(ii)

Find all 2×22 \times 2 matrices A1A_{1}, A2A_{2}, A3A_{3} over C\mathbb {C} such that

tr⁡(A1)=tr⁡(A2)=tr⁡(A3)=0,tr⁡(A1A2)=tr⁡(A2A3)=tr⁡(A3A1)=0. \operatorname {tr}(A_{1}) = \operatorname {tr}(A_{2}) = \operatorname {tr}(A_{3}) = 0, \quad \operatorname {tr}(A_{1}A_{2}) = \operatorname {tr}(A_{2}A_{3}) = \operatorname {tr}(A_{3}A_{1}) = 0.
(iii)

Find all 3×33 \times 3 matrices B1B_{1}, B2B_{2}, B3B_{3} over C\mathbb {C} such that

tr⁡(B1)=tr⁡(B2)=tr⁡(B3)=0,tr⁡(B1B2)=tr⁡(B2B3)=tr⁡(B3B1)=0. \operatorname {tr}(B_{1}) = \operatorname {tr}(B_{2}) = \operatorname {tr}(B_{3}) = 0, \quad \operatorname {tr}(B_{1}B_{2}) = \operatorname {tr}(B_{2}B_{3}) = \operatorname {tr}(B_{3}B_{1}) = 0.
(iv)

Find all 2×22 \times 2 matrices CC such that det⁡(C)=1\operatorname {det}\left(C\right) = 1, tr⁡(C)=0\operatorname {tr}(C) = 0. Do these matrices form a group under matrix multiplication?

Problem 4.2.24

Let AA, BB be n×nn \times n matrices over C\mathbb {C}. Assume that BB is invertible. Show that det⁡(In+BAB−1)=det⁡(In+A)\operatorname {det}\left(I_{n}+BAB^{-1}\right) = \operatorname {det}\left(I_{n}+A\right).

?
Problem 4.2.25

An n×nn \times n matrix AA is called idempotent if A2=AA^{2} = A. Show that rk⁡(A)=tr⁡(A)\operatorname {rk}\left(A\right) = \operatorname {tr}(A).

?
Problem 4.2.26

Let AA be a real symmetric and positive definite n×nn \times n matrix. Then

ln⁡(det⁡(A))≤tr⁡(A)−n. \ln (\operatorname {det}\left(A\right)) \leq \operatorname {tr}(A) - n.

For the identity matrix InI_{n} we obtain equality. Show that for the matrix

A=12(3−1−13) A = \frac{1}{2} \begin{pmatrix} 3 & -1 \\ -1 & 3 \end{pmatrix}

we have n=2n=2, det⁡(A)=2\operatorname {det}\left(A\right) = 2, tr⁡(A)=3\operatorname {tr}(A) = 3 and ln⁡(2)<1\ln (2) < 1.

?
Problem 4.2.27

Let AA, BB, CC be n×nn \times n matrices over C\mathbb {C}. Let a→j\overrightarrow {a}_{j}, b→j\overrightarrow {b}_{j}, c→j\overrightarrow {c}_{j} (j=1,…,nj=1,\ldots ,n) be the jj-th column of AA, BB, CC, respectively. Show that if for some k∈{1,…,n}k \in \left\{ 1,\ldots ,n\right\}

c→k=a→k+b→k \overrightarrow {c}_{k} = \overrightarrow {a}_{k} + \overrightarrow {b}_{k}

and

c→j=a→j=b→j,j=1,…,k−1,k+1,…,n \overrightarrow {c}_{j} = \overrightarrow {a}_{j} = \overrightarrow {b}_{j}, \qquad j=1,\ldots ,k-1,k+1,\ldots ,n

then det⁡(C)=det⁡(A)+det⁡(B)\operatorname {det}\left(C\right) = \operatorname {det}\left(A\right)+\operatorname {det}\left(B\right).

?
Problem 4.2.28

Consider the Legendre polynomials pjp_{j} (j=0,1,…j=0,1,\ldots) with

p0(x)=1,p1(x)=x,p2(x)=12(3x2−1), p_{0}(x) = 1, \qquad p_{1}(x) = x, \qquad p_{2}(x) = \frac{1}{2}(3x^{2}-1), p3(x)=12(5x3−3x),p4(x)=18(35x4−30x2+3). p_{3}(x) = \frac{1}{2}(5x^{3}-3x), \qquad p_{4}(x) = \frac{1}{8}(35x^{4}-30x^{2}+3).

Show that

det⁡(p0(x)p1(x)p2(x)p1(x)p2(x)p3(x)p2(x)p3(x)p4(x))=(1−x2)3(p0(0)0p2(0)0p2(0)0p2(0)0p4(0)). \operatorname {det}\left(\begin{array}{lll} p_{0}(x) & p_{1}(x) & p_{2}(x) \\ p_{1}(x) & p_{2}(x) & p_{3}(x) \\ p_{2}(x) & p_{3}(x) & p_{4}(x) \end{array}\right) = (1-x^{2})^{3} \begin{pmatrix} p_{0}(0) & 0 & p_{2}(0) \\ 0 & p_{2}(0) & 0 \\ p_{2}(0) & 0 & p_{4}(0) \end{pmatrix}.

Note that (xx omitted)

det⁡(p0p1p2p1p2p3p2p3p4) \operatorname {det}\left(\begin{array}{lll} p_{0} & p_{1} & p_{2} \\ p_{1} & p_{2} & p_{3} \\ p_{2} & p_{3} & p_{4} \end{array}\right) =p0(p2p4−(p3)2)−p1(p1p4−p2p3+p2(p1p3−(p2)2) = p_{0}(p_{2}p_{4}-(p_{3})^{2}) - p_{1}(p_{1}p_{4}-p_{2}p_{3} + p_{2}(p_{1}p_{3}-(p_{2})^{2})

and p0(0)=1p_{0}(0) = 1, p1(0)=0p_{1}(0) = 0, p2(0)=−1/2p_{2}(0) = -1/2, p3(0)=0p_{3}(0) = 0, p4(0)=3/8p_{4}(0) = 3/8.

?
Problem 4.2.29

Let AA be an n×nn \times n diagonal matrix over C\mathbb {C}. Let BB be an n×nn \times n matrix over C\mathbb {C} with bjj=0b_{jj} = 0 for all j=1,…,nj=1,\ldots ,n. Can we conclude that all diagonal elements of the commutator [A,B][A,B] are 00?

?
Problem 4.2.30

Let AA, BB be n×nn \times n matrices over R\mathbb {R} with det⁡(A)=1\operatorname {det}\left(A\right) = 1 and det⁡(B)=1\operatorname {det}\left(B\right) = 1. This means AA, BB are elements of the Lie group SL(n,R)SL(n,\mathbb {R}). Can we conclude that

tr⁡(AB)+tr⁡(AB−1)=tr⁡(A)tr⁡(B)? \operatorname {tr}(AB) + \operatorname {tr}(AB^{-1}) = \operatorname {tr}(A)\operatorname {tr}(B)?
?
Problem 4.2.31

Let HH be a hermitian n×nn \times n matrix. Show that det⁡(H+iIn)≠0\operatorname {det}\left(H+iI_{n}\right) \neq 0.

?
Problem 4.2.32

Let AA be an 2×22 \times 2 matrix over R\mathbb {R}. Calculate r=tr⁡(A2)−(tr⁡(A))2r = \operatorname {tr}(A^{2}) - (\operatorname {tr}(A))^{2}. Show that the conditions on ajka_{jk} such that r=0r=0 are given by a12a21=a11a22a_{12}a_{21} = a_{11}a_{22}, i.e. det⁡(A)=0\operatorname {det}\left(A\right) = 0.

?
Problem 4.2.33

Consider a triangle embedded in R3\mathbb {R}^{3}. Let v→j=(xj,yj,zj)\overrightarrow {v}_{j} = (x_{j}, y_{j}, z_{j}) (j=1,2,3j=1,2,3) be the coordinates of the vertices. Then the area AA of the triangle is given by

A=12∥(v→2−v→1)×(v→1−v→3)∥=12∥(v→3−v→1)×(v→3−v→2)∥ A = \frac{1}{2}\left\| (\overrightarrow {v}_{2}-\overrightarrow {v}_{1}) \times (\overrightarrow {v}_{1}-\overrightarrow {v}_{3})\right\| = \frac{1}{2}\left\| (\overrightarrow {v}_{3}-\overrightarrow {v}_{1}) \times (\overrightarrow {v}_{3}-\overrightarrow {v}_{2})\right\|

where ×\times denotes the vector product and ∥⋅∥\left\| \cdot \right\| denotes the Euclidean norm. The area AA of the triangle can also be found via

12(D(x1y11x2y21x3y31))2+(D(y1z11y2z21y3z31))2+(D(z1x11z2x21z3x31))2 \frac{1}{2}\sqrt{\left(D\begin{pmatrix} x_{1} & y_{1} & 1 \\ x_{2} & y_{2} & 1 \\ x_{3} & y_{3} & 1 \end{pmatrix}\right)^{2} + \left(D\begin{pmatrix} y_{1} & z_{1} & 1 \\ y_{2} & z_{2} & 1 \\ y_{3} & z_{3} & 1 \end{pmatrix}\right)^{2} + \left(D\begin{pmatrix} z_{1} & x_{1} & 1 \\ z_{2} & x_{2} & 1 \\ z_{3} & x_{3} & 1 \end{pmatrix}\right)^{2}}

where DD denotes the determinant. Consider

v→1=(1,0,0),v→2=(0,1,0),v→3=(0,0,1). \overrightarrow {v}_{1} = (1,0,0), \qquad \overrightarrow {v}_{2} = (0,1,0), \qquad \overrightarrow {v}_{3} = (0,0,1).

Find the area of the triangle using both expressions. Discuss. The triangle could be one of the faces of a tetrahedron.

?
Problem 4.2.34

Consider the 3×33 \times 3 matrix MM with entries

(M)jk=xkj−1,j,k=1,2,3. (M)_{jk} = x_{k}^{j-1}, \qquad j,k = 1,2,3.

Find the determinant of this matrix.

?
Problem 4.2.35

The permanent and the determinant of an n×nn \times n matrix MM over C\mathbb {C} are respectively defined as

perm⁡(M):=∑π∈Sn(∏j=1nMj,π(j)),det⁡(M):=∑π∈Sn(−1)sgn⁡(π)(∏j=1nMj,π(j)) \operatorname {perm}(M) := \sum _{\pi \in S_{n}} \left(\prod _{j=1}^{n} M_{j,\pi (j)}\right), \qquad \operatorname {det}\left(M\right) := \sum _{\pi \in S_{n}} (-1)^{\operatorname {sgn}(\pi )} \left(\prod _{j=1}^{n} M_{j,\pi (j)}\right)

where SnS_{n} denotes the symmetric group on a set of nn symbols. For an n×nn \times n matrix AA and an m×mm \times m matrix BB we know that

det⁡(A⊗B)≡(det⁡(A))m(det⁡(B))n. \operatorname {det}\left(A \otimes B\right) \equiv (\operatorname {det}\left(A\right))^{m}(\operatorname {det}\left(B\right))^{n}.

Study perm⁡(A⊗B)\operatorname {perm}(A \otimes B).

?
Problem 4.2.36
?
(i)

Let AA, BB be positive definite n×nn \times n matrices. Show that tr⁡(A+B)≥tr⁡(A)\operatorname {tr}(A+B) \geq \operatorname {tr}(A).

(ii)

Let CC, DD be n×nn \times n matrices over C\mathbb {C}. Is tr⁡(CC∗DD∗)≤tr⁡(CC∗)tr⁡(DD∗)\operatorname {tr}(CC^{*}DD^{*}) \leq \operatorname {tr}(CC^{*})\operatorname {tr}(DD^{*})?

Problem 4.2.37

Let AA, BB, CC, DD be n×nn \times n matrices. Assume that DD is invertible. Consider the (2n)×(2n)(2n) \times (2n) matrix

M=(ABCD). M = \begin{pmatrix} A & B \\ C & D \end{pmatrix}.

Show that det⁡(M)=det⁡(AD−BD−1CD)\operatorname {det}\left(M\right) = \operatorname {det}\left(AD-BD^{-1}CD\right) using the identity

(ABCD)(In0n−D−1CIn)≡(A−BD−1CB0nD). \begin{pmatrix} A & B \\ C & D \end{pmatrix} \begin{pmatrix} I_{n} & 0_{n} \\ -D^{-1}C & I_{n} \end{pmatrix} \equiv \begin{pmatrix} A-BD^{-1}C & B \\ 0_{n} & D \end{pmatrix}.
?
Problem 4.2.38

Let μ∈R\mu \in \mathbb {R} and AA be an 2×22 \times 2 matrix over R\mathbb {R}. Show that the determinant of the 4×44 \times 4 matrix

(−μI2AAT−μI2) \begin{pmatrix} -\mu I_{2} & A \\ A^{T} & -\mu I_{2} \end{pmatrix}

only contains even powers of μ\mu.

?
Problem 4.2.39

Consider the invertible 3×33 \times 3 matrix AA

A=(1/201/21/20−1/2010)  ⇒  A−1=(1/21/200011/2−1/20). A = \begin{pmatrix} 1/\sqrt{2} & 0 & 1/\sqrt{2} \\ 1/\sqrt{2} & 0 & -1/\sqrt{2} \\ 0 & 1 & 0 \end{pmatrix} \; \Rightarrow \; A^{-1} = \begin{pmatrix} 1/\sqrt{2} & 1/\sqrt{2} & 0 \\ 0 & 0 & 1 \\ 1/\sqrt{2} & -1/\sqrt{2} & 0 \end{pmatrix}.

Find a permutation matrix PP such that PA−1P−1=APA^{-1}P^{-1} = A.

?
Problem 4.2.40

Let x,ϵ∈Rx, \epsilon \in \mathbb {R}. Show that the determinant of the symmetric n×nn \times n matrix

A(x,ϵ)=(x+ϵxxxx+ϵx⋱xxx+ϵ) A(x,\epsilon ) = \begin{pmatrix} x+\epsilon & x & & x \\ x & x+\epsilon & & x \\ & & \ddots & \\ x & x & & x+\epsilon \end{pmatrix}

is given by det⁡(A(x,ϵ))=ϵn−1(nx+ϵ)\operatorname {det}\left(A(x,\epsilon )\right) = \epsilon^{n-1}(nx+\epsilon ).

?
Problem 4.2.41

Show that the determinant of the 2n×2n2n \times 2n matrix

Ω=(0nIn−In0n)≡(01−10)⊗In \Omega = \begin{pmatrix} 0_{n} & I_{n} \\ -I_{n} & 0_{n} \end{pmatrix} \equiv \begin{pmatrix} 0 & 1 \\ -1 & 0 \end{pmatrix} \otimes I_{n}

is given by det⁡(Ω)=(det⁡(01−10))n(det⁡(In))2=1\operatorname {det}\left(\Omega \right) = \left(\operatorname {det}\left(\begin{array}{ll} 0 & 1 \\ -1 & 0 \end{array}\right)\right)^{n} (\operatorname {det}\left(I_{n}\right))^{2} = 1.

?
Problem 4.2.42

Given a 2×2×22 \times 2 \times 2 hypermatrix A=(ajkℓ)A = (a_{jk\ell }), j,k,ℓ=0,1j,k,\ell = 0,1 and the 2×22 \times 2 matrix

S=(s00s01s10s11). S = \begin{pmatrix} s_{00} & s_{01} \\ s_{10} & s_{11} \end{pmatrix}.

The multiplication ASAS which is again a 2×22 \times 2 hypermatrix is defined by

(AS)jkℓ:=∑r=01ajkrsrℓ. (AS)_{jk\ell } := \sum _{r=0}^{1} a_{jkr}s_{r\ell }.

Assume that det⁡(S)=1\operatorname {det}\left(S\right) = 1, i.e. S∈SL(2,C)S \in SL(2,\mathbb {C}). Show that Det⁡(AS)=Det⁡(A)\operatorname {Det}(AS) = \operatorname {Det}(A). This is a typical problem to apply computer algebra.

?
Problem 4.2.43

The Levi-Civita symbol (also called completely antisymmetric constant tensor) is defined by

ϵj1,j2,…,jn:={+1,if j1,j2,…,jn is an even permutation of 12…n−1,if j1,j2,…,jn is an odd permutation of 12…n0,otherwise \epsilon _{j_{1},j_{2},\ldots ,j_{n}} := \begin{cases} +1, & \text{if } j_{1},j_{2},\ldots ,j_{n} \text{ is an even permutation of } 12\ldots n \\ -1, & \text{if } j_{1},j_{2},\ldots ,j_{n} \text{ is an odd permutation of } 12\ldots n \\ 0, & \text{otherwise} \end{cases}

Let δjk\delta_{jk} be the Kronecker delta. Show that

ϵj1,j2,…,jnϵk1,k2,…,kn=det⁡(δj1k1δj2k1δjnk1δj1k2δj2k2δjnk2⋮⋮⋮δj1knδj2kn…δjnkn). \epsilon _{j_{1},j_{2},\ldots ,j_{n}} \epsilon _{k_{1},k_{2},\ldots ,k_{n}} = \operatorname {det}\left(\begin{array}{llll} \delta _{j_{1}k_{1}} & \delta _{j_{2}k_{1}} & & \delta _{j_{n}k_{1}} \\ \delta _{j_{1}k_{2}} & \delta _{j_{2}k_{2}} & & \delta _{j_{n}k_{2}} \\ \vdots & \vdots & & \vdots \\ \delta _{j_{1}k_{n}} & \delta _{j_{2}k_{n}} & \ldots & \delta _{j_{n}k_{n}} \end{array}\right).
?