Scribd
Upload
21 views

Tutorial 8 Soln

This document provides solutions to problems from a tutorial on linear algebra. It addresses topics like row space, null space, and rank of matrices. Key steps and counterexamples are provided to justify answers. Matrix transformations and their effects on properties like rank are discussed.

Uploaded by

stayhungry0105
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
21 views

Tutorial 8 Soln

This document provides solutions to problems from a tutorial on linear algebra. It addresses topics like row space, null space, and rank of matrices. Key steps and counterexamples are provided to justify answers. Matrix transformations and their effects on properties like rank are discussed.

Uploaded by

stayhungry0105
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

THE UNIVERSITY OF HONG KONG

DEPARTMENT OF MATHEMATICS

MATH2101 Linear Algebra I


Tutorial 8 solution

1. (a) For all x ∈ Row(CA), x = yT (CA) for some y ∈ Rm . Then x = zT A for some z = C T y ∈ Rn
and hence x ∈ Row A. Thus Row(CA) ⊆ Row A.
(b) Using (a) for C = B, we have Row(BA) ⊆ Row A.
On the other hand, as B is invertible, using (a) for C = B −1 ,

Row A = Row(B −1 BA) ⊆ Row(BA).

Thus Row(BA) = Row A.

2. The correct answer: (a),(b),(d).


(a) Note that A and B can be transformed into the same reduced row echelon form R after
applying some elementary row operations. To be more precise, there exist some elementary
matrices E1 , . . . , Ek such that

R = Ek . . . E1 B = Ek . . . E1 EA.

(b) It follows from 1(b).


1 0 0 1 0 0
(c) A counter example: A = [ ] and E = [ ]. Then B = [ ] and hence
0 0 1 0 1 0

Col A = Span {e1 } ≠ Span {e2 } = Col B.

(d) As E is invertible, EAx = 0 if and only if Ax = 0. So Null B = Null EA = Null A.

3. By the assumption, Null A = Span{ei1 , . . . , eik } for 1 ≤ i1 < i2 < . . . < ik ≤ n. For m = 1, . . . , k,
the im -th column of A is Aeim = 0 and hence A contains at least k zero columns.
On the other hand, A cannot contains more than k zero columns, otherwise there exists
ℓ ∈ {1, 2, . . . , n}/{i1 , i2 , . . . , ik } such that the ℓ-th column of A is a zero vector, that is, Aeℓ = 0,
which yields
eℓ ∈ Null A = Span{ei1 , . . . , eik }.
This is a contradiction.
Therefore, A has exactly k zero columns.

4. The correct answer: (a).


1 −1
Note that v = u1 + 2u2 = u1 − (−2u2 ) = −(−u2 ) + (u1 + u2 ). So [v]B2 = [ ] and [v]B3 = [ ].
−1 1

1
1 1 T
(a) [v]B2 + [u1 ]B1 = [ ] + [ ] = [2 −1] .
−1 0

1
(b) Span{[v]B2 , [v]B3 } = Span {[ ]} ≠ R2 .
−1

1 −1
(c) {[v]B1 , [v]B3 } = {[ ] , [ ]} is linearly independent.
2 1

1 3
(d) {[v]B2 + [3u2 ]B1 , 3[v]B1 } = {[ ] , [ ]} is linearly dependent.
2 6

0 1
5. (a) For example, let A = [ ]. Then
0 0

0 b
Row A = {[ ] ∶ a ∈ R} = Span {e2 } ≠ Span {e1 } = {[ ] ∶ b ∈ R} = Col A,
a 0

where row vectors are identified as column vectors in the natural sense.
0 1
(b) For example, let A = [ ]. Then
0 0

a b
Col A = {[ ] ∶ a ∈ R} = Span {e1 } = {[ ] ∶ b ∈ R} = Null A.
0 0

6. (a) True. Note that BA and A are both m × n matrices. It follows from that 1(b) that

rankA=dim(Row A) = dim(Row(BA)) =rank(BA).

Remark. Alternatively, we may consider the nullity to prove the result:

rankA + dim(Null A) = n = rank(BA) + dim(Null(BA)) (1)

As B is invertible, Ax = 0 ⇐⇒ BAx = 0 and hence Null A = Null(BA). Thus

dim(Null A)=dim(Null(BA)).

It follows from (1) that rankA=rank(BA).


(b) False. Consider m = n = 2. Let A = B = O. Then rank(BA)=0=rankA. But B is not
invertible.

You might also like