Tutorial Letter 101/3/2017: Linear Algebra

MAT2611/101/3/2017
Tutorial letter 101/3/2017
LINEAR ALGEBRA
MAT2611
Semesters 1 & 2
Department of Mathematical Sciences
IMPORTANT INFORMATION:
This tutorial letter contains important information about your
module.
BARCODE
university
Define tomorrow. of south africa
CONTENTS
Page
1 INTRODUCTION ..................................................................................................................5
2 PURPOSE AND OUTCOMES FOR THE MODULE ............................................................5
2.1 Purpose ................................................................................................................................5
2.2 Outcomes .............................................................................................................................5
3 LECTURER(S) AND CONTACT DETAILS ..........................................................................6
3.1 Lecturer(s) ............................................................................................................................6
3.2 Department ..........................................................................................................................6
3.3 University..............................................................................................................................7
4 RESOURCES.......................................................................................................................7
4.1 Prescribed books..................................................................................................................7
4.2 Recommended books ..........................................................................................................7
4.3 Electronic reserves (e-Reserves) .........................................................................................8
4.4 Library services and resources information .........................................................................8
5 STUDENT SUPPORT SERVICES .......................................................................................9
6 STUDY PLAN.......................................................................................................................9
7 PRACTICAL WORK AND WORK INTEGRATED LEARNING............................................9
8 ASSESSMENT.....................................................................................................................9
8.1 Assessment criteria ..............................................................................................................9
8.2 Assessment plan ................................................................................................................11
8.3 Assignment numbers..........................................................................................................11
8.3.1 General assignment numbers ............................................................................................11
8.3.2 Unique assignment numbers..............................................................................................11
8.3.3 Assignment due dates........................................................................................................12
8.4 Submission of assignments................................................................................................12
8.5 The assignments ................................................................................................................12
8.6 Other assessment methods ...............................................................................................13
8.7 The examination .................................................................................................................13
9 FREQUENTLY ASKED QUESTIONS................................................................................13
10 IN CLOSING ......................................................................................................................13
ADDENDUM A: ASSIGNMENTS – FIRST SEMESTER ..............................................................14
2
MAT2611/101/3/2017
ADDENDUM B: ASSIGNMENTS – SECOND SEMESTER .........................................................22

ADDENDUM C: EXAM INFORMATION SHEET ..........................................................................31
ADDENDUM D: USEFUL COMPUTER SOFTWARE ..................................................................38
ADDENDUM E: ELEMENTARY LINEAR ALGEBRA USING MAXIMA ......................................39
E.1 The linearalgebra and eigen packages ....................................................................39
E.2 Matrices..............................................................................................................................39
E.3 Eigenvalues and eigenvectors............................................................................................40
E.4 Rank, nullity, columnspace and nullspace ..........................................................................41
E.5 Matrix inverse .....................................................................................................................41
E.6 Gram-Schmidt algorithm ....................................................................................................41
ADDENDUM F: Example questions ...........................................................................................44
F.1 Previous assignment questions..........................................................................................44
Questions ...........................................................................................................................44
Solutions.............................................................................................................................57
F.2 Previous multiple choice questions ..................................................................................130
Questions .........................................................................................................................130
Solutions...........................................................................................................................146
F.3 2016 Semester 1: Exam...................................................................................................175
Question paper.................................................................................................................175
Solution ............................................................................................................................181
F.4 2015 Semester 1: Exam...................................................................................................192
Question paper.................................................................................................................192
Solution ............................................................................................................................197
F.5 2014 Semester 1: Exam...................................................................................................208
Question paper.................................................................................................................208
Solution ............................................................................................................................214
F.6 2013 Semester 1: Exam...................................................................................................225
Question paper.................................................................................................................225
Solution ............................................................................................................................228
F.7 2012 Semester 1: Exam...................................................................................................236
Question paper.................................................................................................................236
Solution ............................................................................................................................238
3
F.8 2012 Semester 2: Exam...................................................................................................245
Question paper.................................................................................................................245
Solution ............................................................................................................................247
F.9 2010 Semester 2: Exam...................................................................................................255
Question paper.................................................................................................................255
Solution ............................................................................................................................258
4
MAT2611/101/3/2017
1 INTRODUCTION
Dear Student
Welcome to the MAT2611 module in the Department of Mathematical Sciences at Unisa. We trust
that you will find this module both interesting and rewarding.
Some of this tutorial matter may not be available when you register. Tutorial matter that is not
available when you register will be posted to you as soon as possible, but is also available on
myUnisa.
myUnisa
You must be registered on myUnisa (http://my.unisa.ac.za) to be able to submit assignments

online, gain access to the library functions and various learning resources, download study ma-
terial, “chat” to your lecturers and fellow students about your studies and the challenges you en-
counter, and participate in online discussion forums. myUnisa provides additional opportunities to
take part in activities and discussions of relevance to your module topics, assignments, marks and
examinations.
Tutorial matter
A tutorial letter is our way of communicating with you about teaching, learning and assessment.
You will receive a number of tutorial letters during the course of the module. This particular tutorial
letter contains important information about the scheme of work, resources and assignments for this
module as well as the admission requirements for the examination. We urge you to read this and
subsequent tutorial letters carefully and to keep it at hand when working through the study material,
preparing and submitting the assignments, preparing for the examination and addressing queries
that you may have about the course (course content, textbook, worked examples and exercises,
theorems and their applications in your assignments, tutorial and textbook problems, etc.) to your
MAT2611 lecturers.
2 PURPOSE AND OUTCOMES FOR THE MODULE

2.1 Purpose
This module is a direct continuation of MAT1503. It will be useful to students interested in develop-
ing their Linear Algebra techniques and skills in solving problems in the mathematical sciences.
2.2 Outcomes
To understand, compute and apply the following linear algebra concepts:
2.2.1 Vector spaces

(Anton & Rorres, sections 4.1 – 4.5), (Lay, sections 4.1 – 4.5).
2.2.2 Rank of a matrix

(Anton & Rorres, sections 4.7 – 4.8), (Lay, section 4.6).
5
2.2.3 Change of basis
(Anton & Rorres, section 4.6), (Lay, section 4.7).
2.2.4 Eigenvalues and eigenvectors

(Anton & Rorres, section 5.1), (Lay, sections 5.1 – 5.2).
2.2.5 Diagonalisation of matrices

2.2.6 Inner products and orthogonality

(Anton & Rorres, sections 6.1 – 6.2), (Lay, sections 6.1 – 6.3).
2.2.7 Gram-Schmidt algorithm

2.2.8 Orthogonal diagonalisation of symmetric matrices

(Anton & Rorres, sections 7.1 – 7.2), (Lay, section 7.1).
2.2.9 Linear transformations

(Anton & Rorres, chapter 8), (Lay, section 4.2 and section 5.4).
3 LECTURER(S) AND CONTACT DETAILS

3.1 Lecturer(s)
The contact details for the lecturer responsible for this module is
Postal address: The MAT2611 Lecturers

Department of Mathematical Sciences
Private Bag X6
Florida
1709
South Africa
Additional contact details for the module lecturers will be provided in a subsequent tutorial letter.
All queries that are not of a purely administrative nature but are about the content of this module
should be directed to your lecturer(s). Tutorial letter 301 will provide additional contact details for
your lecturer. Please have your study material with you when you contact your lecturer by tele-
phone. If you are unable to reach us, leave a message with the departmental secretary. Provide
your name, the time of the telephone call and contact details. If you have problems with questions
that you are unable to solve, please send your own attempts so that the lecturers can determine
where the fault lies.
Please note: Letters to lecturers may not be enclosed with or inserted into assignments.
3.2 Department
The contact details for the Department of Mathematical Sciences are:
6
MAT2611/101/3/2017
Departmental Secretary: (011) 670 9147 (SA) +27 11 670 9147 (International)
3.3 University
If you need to contact the University about matters not related to the content of this module, please
consult the publication Study @ Unisa that you received with your study material. This booklet
contains information on how to contact the University (e.g. to whom you can write for different
queries, important telephone and fax numbers, addresses and details of the times certain facilities
are open). Always have your student number at hand when you contact the University.
4 RESOURCES
4.1 Prescribed books
Prescribed books can be obtained from the University’s official booksellers. If you have difficulty
locating your book(s) at these booksellers, please contact the Prescribed Books Section at (012)
429 4152 or e-mail vospresc@unisa.ac.za.
Your prescribed textbook for this module is:
Title: Elementary Linear Algebra with Supplemental Applications

Authors: Howard Anton and Chris Rorres
Edition: 11th Edition, International Student Version
Publishers: Wiley
ISBN: 978-1-118-67745-2
However, you may wish to use your copy of
Title: Linear Algebra and Its Applications

Author: David C. Lay
Edition: Pearson New International Edition, 4th edition
Publishers: Pearson
ISBN: 9781292020556
Students with the textbook by Lay will be accommodated.
Please buy the textbook as soon as possible since you have to study from it directly – you cannot
do this module without the prescribed textbook.
4.2 Recommended books

The book “Linear Algebra” by Jim Hefferon is available for free from
http://joshua.smcvt.edu/linearalgebra/
with answers to exercises available from the same web site. The concepts are arranged differently
to the prescribed book. The relevant chapters and sections are: chapter 2, chapter 3 I-III and V.
Some of the terminology is different to the prescribed book.
7
The book “A First Course in Linear Algebra” by Robert A. Beezer is a free and interactive online
book available at
http://linear.ups.edu/
and also has multiple PDF versions available for download. The relevant chapters are “Vectors”,
“Matrices” - “Column and Row Spaces”, “Vector Spaces”, “ Eigenvalues”, “Linear Transformations”
and “Representations”. Please note that this book assumes that vector spaces are over the field
of complex numbers, while the prescribed text book considers only the real numbers.
Finally, the “Book of Proof” (Second Edition) by Richard Hammack, Part I, Chapter 1 (Sets) is
recommended for students who need to revise basic set theory and notation. The entire book is
available for free from
http://www.people.vcu.edu/~rhammack/BookOfProof/index.html
4.3 Electronic reserves (e-Reserves)

There are no e-Reserves for this module.
4.4 Library services and resources information

For brief information go to:
http://www.unisa.ac.za/brochures/studies
For more detailed information, go to the Unisa website: http://www.unisa.ac.za/, click on Li-
brary. For research support and services of Personal Librarians, go to:
http://www.unisa.ac.za/Default.asp?Cmd=ViewContent&ContentID=7102
The Library has compiled numerous library guides:
• find recommended reading in the print collection and e-reserves

- http://libguides.unisa.ac.za/request/undergrad
• request material
- http://libguides.unisa.ac.za/request/request
• postgraduate information services

- http://libguides.unisa.ac.za/request/postgrad
• finding , obtaining and using library resources and tools to assist in doing research
- http://libguides.unisa.ac.za/Research_Skills
• how to contact the Library/find us on social media/frequently asked questions

- http://libguides.unisa.ac.za/ask
8
MAT2611/101/3/2017
5 STUDENT SUPPORT SERVICES

For information on the various student support services available at Unisa (e.g. student counseling,
tutorial classes, language support), please consult the publication Study @ Unisa that you received
with your study material.
6 STUDY PLAN
The following table provides an outline of the outcomes and ideal dates of completion, and other
study activities.
Semester 1 Semester 2
Outcomes 2.2.1–2.2.3 to be achieved by 17 February 2017 3 August 2017
Outcomes 2.2.3–2.2.5 to be achieved by 16 March 2017 31 August 2017
Outcomes 2.2.6–2.2.9 to be achieved by 13 April 2017 28 September 2017
See the brochure Study @ Unisa for general time management and planning skills.
7 PRACTICAL WORK AND WORK INTEGRATED LEARNING

There are no practicals for this module.
8 ASSESSMENT
8.1 Assessment criteria
Specific outcome 1: Understand and apply the definition of a general real vector space, along
with the concepts of subspace, linear independence, basis and dimension, row space column
space and null space, rank and nullity.
Assessment criteria
You must be able to do the following.
• Decide, with reasons, whether a given set with two given operations defines a vector space.
• Decide, with reasons, whether a given subset of a vector space defines a subspace.
• Find the span of a given set of vectors. Show that a given set of vectors do/do not span a
given space, with reasons.
• Test a given set of vectors for linear dependence/independence.
• Find a basis for a given vector space. Find a basis for the span of a given set of vectors.
Determine whether or not a given set of vectors forms a basis for a given vector space.
• Find for a given matrix the row space/column space and null space.
9
• Find, for a given linear system, the general solution.
• Find the rank and nullity of a given matrix.
Specific outcome 2: Understand and be able to apply the basis concepts of inner product
spaces.
Assessment criteria
• Calculate inner products in cases other than the dot product.
• Use the length, angle and distance formulas for arbitrary inner products.
• Test vectors for orthogonality.
• Find orthogonal complements of subspaces.
• Test sets of vectors for orthogonality/orthonormality.
• Use the Gram-Schmidt process to change a basis to an orthogonal/orthonormal one.
• Find the transition matrix between two different bases.
• Find the coordinate vector of a vector with respect to a new basis.
• Decide whether or not a matrix is orthogonal.
Specific outcome 3: Understand and be able to apply the basis concepts of eigenvalues and
eigenvectors.
Assessment criteria
• Test whether a given scalar/vector pair is an eigenvalue–eigenvector pair of a matrix.
• Find the eigenvalues and eigenvectors of a matrix.
• Determine whether or not a given matrix is diagonalisable, giving reasons.
• Find, for a diagonalisable matrix, a diagonalising matrix.
• Determine whether or not a given matrix is orthogonally diagonalisable, giving reasons.
• Find, for an orthogonally diagonalisable matrix, an orthogonal diagonalising matrix.
10
MAT2611/101/3/2017
Specific outcome 4: Understand and be able to apply the concept of linear transformation.
Assessment criteria
The student must be able to:
• Decide, with reasons, whether a given operation on vector space is a linear transformation or
not.
• Find the kernel and range of a linear transformation.
• Find the rank and nullity of a linear transformation.
• Determine whether a given linear transformation is one-to-one and/or onto.
• Find, in those cases where is is possible, the inverse of a linear transformation.
• Find the matrix of a linear transformation with respect to a given basis.
• Find the matrices of compositions of transformations and inverse transformations with respect
to a given basis.
• Find the matrix of a linear transformation with respect to a basis, given the matrix with respect
to a different basis.
• Find the eigenvalues of a linear operator.
• Decide if a given linear transformation is an isomorphism or not, with reasons.
8.2 Assessment plan

A final mark of at least 50% is required to pass the module. If a student does not pass the module
then a final mark of at least 40% is required to permit the student access to the supplementary
examination. The final mark is composed as follows:
Year mark Final mark

Assignment 01: 30% −→ Year mark: 20%
Assignment 02: 40% Exam mark: 80%
Assignment 03: 30%
8.3 Assignment numbers

8.3.1 General assignment numbers
The assignments for this module are Assignment 01, Assignment 02, etc.
8.3.2 Unique assignment numbers

Please note that each assignment has a unique assignment number which must be written on the
cover of your assignment.
11
8.3.3 Assignment due dates
The dates for the submission of the assignments are:
Semester 1
Assignment 01: Friday, 24 February 2017

Assignment 02: Thursday, 23 March 2017
Assignment 03: Thursday, 20 April 2017
Semester 2
Assignment 01: Thursday, 10 August 2017

Assignment 02: Thursday, 7 September 2017
Assignment 03: Thursday, 5 October 2017
8.4 Submission of assignments

You may submit written assignments either by post or electronically via myUnisa. Assignments
may not be submitted by fax or e-mail.
For detailed information on assignments, please refer to the Study @ Unisa

brochure which you received with your study package.
Please make a copy of your assignment before you submit!
To submit an assignment via myUnisa:
• Go to my Unisa.
• Log in with your student number and password.
• Select the module.
• Click on “Assignments” in the menu on the left-hand side of the screen.
• Click on the assignment number you wish to submit.
• Follow the instructions.
8.5 The assignments

Please make sure that you submit the correct assignments for the 1st semester, 2nd semester or
year module for which you have registered. For each assignment there is a fixed closing date,
the date at which the assignment must reach the University. When appropriate, solutions for each
assignment will be dispatched, as Tutorial Letter 201 (solutions to Assignment 01) and Tutorial
Letter 202 (solutions to Assignment 02) etc., a few days after the closing date. They will also be
12
MAT2611/101/3/2017
made available on my Unisa. Late assignments will not be marked!
Note that at least one assignment must reach us before the due date in order to gain admis-
sion to the examination.
8.6 Other assessment methods

There are no other assessment methods for this module.
8.7 The examination

During the relevant semester, the Examination Section will provide you with information regarding
the examination in general, examination venues, examination dates and examination times. For
general information and requirements as far as examinations are concerned, see the brochure
Study @ Unisa.
Registered for . . . Examination period Supplementary examination period

1st semester module May/June 2017 October/November 2017
2nd semester module October/November 2017 May/June 2018
Year module October/November 2017 January/February 2018
9 FREQUENTLY ASKED QUESTIONS

The Study @ Unisa brochure contains an A–Z guide of the most relevant study information.
10 IN CLOSING
We hope that you will enjoy MAT2611 and we wish you all the best in your studies at Unisa!
13
ADDENDUM A: ASSIGNMENTS – FIRST SEMESTER
ASSIGNMENT 01
Due date: Friday, 24 February 2017
UNIQUE ASSIGNMENT NUMBER: 739590
ONLY FOR SEMESTER 1
This assignment is a multiple choice assignment. Please consult the Study @ Unisa
brochure for information on how to submit your answers for multiple choice assignments.
Question 1
Consider the set

a 1
X := : a ∈ R ⊂ M22
0 −a

a 1 a 1
and the operations for all k, a, b ∈ R, u = ∈ X and v = ∈X
0 −a 0 −a

a 1 ka 1
· : R × X → X, k·u≡k· := ,
0 −a 0 −ka

a 1 b 1 a+b 1
+ : X × X → X, u+v ≡ + :=
0 −a 0 −b 0 −(a + b)
The set X with these definitions of · and + forms a vector space. The zero vector for X is

0 0
1.
0 0

1 0
2.
0 −1

1 0
3.
0 0

0 1
4.
0 0
5. None of the above.
Question 2
Which of the following are subspaces of P2 with the usual operations ?

A. span { 1, x2 }
B. { a + ax : a ∈ R }
14
MAT2611/101/3/2017
a + 1b x : a, b ∈ R, b 6= 0

C.
D. { ax3 : a ∈ R }
Select from the following:
1. All of A, B, C and D.
2. Only A, B, and D.
3. Only A and B.
4. Only B and D.
Question 3
Which of the following sets are linearly independent?

A. { (1, 0), (1, 1), (1, −1) } in R2
B. { (1, 1, 1), (1, −1, 1), (−1, 1, 1) } in R3
C. { 1 + x, x, 2 + 3x } in P2

1 1 1 1
D. , in M22
0 0 0 1
1. Only A and C.
2. Only B.
3. Only D.
4. Only B and D.
Question 4
Which of the following sets are a basis for the following vector subspace of R3 ?
   
 0 1 0 
X = x ∈ R3 : 0 0 1 x = x .
1 0 0
 
         
 0   1   2   0 1 
A.  0 B. 1 C.  2 D. 1 , 0
0 1 2 1 0
       
15
1. Only A and B.
2. Only B and C.
3. Only C and D.
4. Only A and D.
Question 5
Which of the following statements are true:
A. dim(span { (1, 0, 1), (1, 0, −1) }) = 2 in R3
B. dim(span { (1, 0, 0), (−1, 0, 0) }) = 2 in R3
C. dim(span { (1, 1, 1), (1, 1, −1), (1, −1, 1), (−1, 1, 1) }) = 4 in R3
1. All of A, B and C.
2. Only A and B.
3. Only A and C.
4. Only A.
Question 6
 
1 3
Which of the following sets are a basis for the row space of 1 1 ?
3 1

A. 1 3 , 1 1 , 3 1

B. 1 −1 , 0 1

C. 1 −1 , 1 1

D. 1 2 , 2 1
1. Only A.
2. Only B, C and D.
3. Only B and C.
16
MAT2611/101/3/2017
Question 7

1 1 0
Which of the following sets are a basis for the null space of ?
1 −1 0
n T o
A. 0 0 1
n T T o
B. 1 1 0 , 1 −1 0
n T T o
C. 2 0 0 , 1 1 0
n T T o
D. 1 1 , 1 −1
1. Only B.
2. Only D.
3. Only B and C.
4. Only A.
Question 8
Which of the following statements are always true for for all m, n ∈ N and m × n matrices A ?
A. rank(A) = rank(AT )
B. rank(AT ) + nullity(AT ) = m
C. rank(AT ) + nullity(AT ) = n
D. row space(A) = column space(A)
1. Only A and C.
2. Only A and B.
3. Only B.
4. Only C.
– End of assignment –
17
ASSIGNMENT 02
Due date: Thursday, 23 March 2017
Total Marks: 40
ONLY FOR SEMESTER 1
Answer all the questions. Show all your workings.
If you choose to submit via my Unisa, note that only PDF files will be accepted.
Question 1: 20 Marks
Let            
 1
 1 0   1
 1 1 
 1 1 0   1 1  1 
B2 = 1 ,
  ,
0
 
1  and B2 = 1 ,
  ,
1
 
−1 

  
 
0 0 1 1 −1 1
   
be two bases for span(B1 ), where the usual left to right ordering is assumed.
(1.1) Find the transition matrix (change of coordinate/change of basis matrix) PB1 →B2 . (8)
(1.2) Let B3 be a basis for R3 and let the transition matrix from B2 to B3 be given by
 
1 1 0
PB2 →B3 = 1 0 1 .
1 1 1
(a) Find the transition matrix PB1 →B3 . (6)

(b) Use PB2 →B3 to find B3 . (6)
Consider the matrix  

1 0 1 0
0 1 0 0
A=
0
.
0 0 0
1 0 0 1
(2.1) Determine the characteristic equation for A in λ. (4)
(2.2) Find the eigenvalues of A, and their algebraic multiplicities. (4)
(2.3) Find a basis for the eigenspace corresponding to each eigenvalue of A and hence also (12)
the geometric multiplicity of each eigenvalue.
18
MAT2611/101/3/2017
ASSIGNMENT 03
Due date: Thursday, 20 April 2017
ONLY FOR SEMESTER 1
Question 1
Let A be an n × n matrix, x ∈ Rn and λ ∈ R. The equation Ax = λx for x has the unique solution
x = 0 if and only if
1. λ = 0.
2. λ is not an eigenvalue of A.
3. λ = 0 and 0 is an eigenvalue of A.
4. A is invertible.
Question 2
Let A be an n × n matrix with eigenvalue −1, In be the n × n identity matrix and 0n be the n × n
zero matrix. Which of the following are true?
A. (−1)k is an eigenvalue of Ak for all k ∈ N.
B. In + A is singular.
C. In + A = 0n .
D. If x ∈ Rn such that Ax = −x, then x = 0.
1. Only A, B and C.
2. Only A and B.
3. Only A and D.
4. Only B, C and D.
19
Question 3
Which of the following matrices are diagonalizable?

 
1 1 0
1 0 1 2 1 0
A. 0 −1 1. B. . C. . D. .
0 1 0 −1 2 −1
0 0 1
1. Only A.
2. Only B.
3. Only B and C.
4. Only B, C and D.
Question 4
Let A and B be n × n matrices and let In be the n × n identity matrix. Then
1. AB + B T AT is diagonalizable.
2. If A is invertible then A is diagonalizable.
3. If A and B are diagonalizable then A + B is diagonalizable.
4. If λ = 0 is and eigenvalue of A, then A is not diagonalizable.
Question 5
Which one of the following defines an inner product?

1 −1 T
1. hA, Bi = tr AB in M22 .
−1 1
2. ha1 + b1 x + c1 x2 , a2 + b2 x + c2 x2 i = a1 b1 + a2 b2 in P2 .
3. h(x1 , x2 ), (y1 , y2 )i = x1 y1 + 2x2 y2 in R2 .
4. h(x1 , x2 ), (y1 , y2 )i = x1 y1 + 2x2 y2 − 1 in R2 .
20
MAT2611/101/3/2017
Question 6
Which of the following vectors are unit vectors with respect to the inner product
h(x1 , x2 , x3 ), (y1 , y2 , y3 )i = 2x1 y1 + 2x2 y2 + 2x3 y3 in R3 ?
√ √
A. (1, 0, 0) B. (1, 0, 0)/ 2 C. (1, 0, 1)/ 2 D. (1, 1, 0)/2
1. Only A.
2. Only B and D.
3. Only A and C.
Question 7
Which of thefollowing
vectors
are orthogonal to each other with respect to the inner product
1 0 T
hA, Bi = tr A B in M22 ?
0 2

1 1 1 −1 1 1 1 1
A. . B. . C. . D. .
1 −1 1 1 −1 1 1 1
1. All of A, B, C and D are orthogonal to each other.
2. None of A, B, C and D are orthogonal to each other.
3. Only A and B are orthogonal, A and C are orthogonal, B and C are orthogonal.
4. Only A and C are orthogonal, B and C are orthogonal.
Question 8
Consider the vector subspace W = span{ 1 − x, 2x2 } of P2 with the evaluation inner product at 0, 1
and −1 (sample points). Which of the following vectors in P2 lie in the subspace W ⊥ of P2 ?
1. x2 − 1.
2. x2 + x + 1.
3. x.
4. −2x2 + x + 2.
21
ADDENDUM B: ASSIGNMENTS – SECOND SEMESTER
ASSIGNMENT 01
Due date: Thursday, 10 August 2017
ONLY FOR SEMESTER 2
Question 1
Consider the set

a 1
X := : a ∈ R ⊂ M22
0 −a

a 1 a 1
and the operations for all k, a, b ∈ R, u = ∈ X and v = ∈X
0 −a 0 −a

a 1 ka 1
· : R × X → X, k·u≡k· := ,
0 −a 0 −ka

a 1 b 1 a+b 1
+ : X × X → X, u+v ≡ + :=
0 −a 0 −b 0 −(a + b)
The set X with these definitions of · and + forms a vector space. Which one of the following statements
are true in this vector space?

1 1 −1 −1
1. − =
0 −1 0 1

1 1 −1 −1
2. − =
0 −1 0 −1

1 1 1 1
3. − =
0 −1 0 1

1 1 −1 1
4. − =
0 −1 0 1
Question 2
Which of the following are subspaces of P2 with the usual operations ?
A. span { 1, x2 }
22
MAT2611/101/3/2017
B. { 1 + ax : a ∈ R }
C. { a − bx2 : a, b ∈ R }
D. { a : a ∈ R, a ≥ 0 }
1. Only A, B and C.
2. Only A, C and D.
3. Only C and D.
4. Only A and C.
Question 3

A. { (1, 0), (1, 1), (1, −1) } in R2
B. { (1, 1, 1), (1, −1, 1), (2, −3, 2) } in R3
C. { 1 + x, x, 2 + 3x } in P2

1 1 1 −1
D. , in M22
1 1 1 −1
1. Only A, B and C.
2. Only B and C.
3. Only B and D.
4. Only D.
Question 4
Which of the following sets are a basis for the following vector subspace of M22 :

1 0
X = A ∈ M22 : A = .
−1 0

0 0
A.
0 0
23

1 1
B.
1 1

1 1 0 0
C. ,
0 0 1 1

1 1 1 1
D. ,
1 1 0 0
1. Only A and B.
2. Only B and C.
3. Only C and D.
4. Only A and D.
Question 5
A. dim(span { 1 + x2 , 1 − x2 }) = 2 in P2
B. dim(span { x2 , −x2 }) = 2 in P2
C. dim(span { 1 + x + x2 , 1 + x − x2 , 1 − x + x2 , −1 + x + x2 }) = 4 in P2
1. All of A, B, and C.
2. Only A and C.
3. Only A and B.
4. Only A.
Question 6

1 1 3
Which of the following sets are a basis for the column space of ?
3 1 1

1 1 3
A. , ,
3 1 1
24
MAT2611/101/3/2017

1 1
B. ,
1 2

1 1
C. ,
3 1

1 1
D. ,
0 1
2. Only B, C and D.
3. Only A.
4. Only B and C.
Question 7

1 1 −1
0 −1 1
n T o
A. 0 1 1
n T T o
B. 0 1 1 , 2 −1 1
n T T o
C. 1 1 −1 , 0 −1 1
n T T o
D. 1 0 , 1 −1
1. Only A.
2. Only C.
3. Only B.
4. Only A.
25
Question 8
Which of the following statements are always true for for all m, n ∈ N and m × n matrices A ?
A. rank(A) = rank(AT )
B. rank(AT ) + nullity(AT ) = m
C. rank(AT ) + nullity(AT ) = n
1. Only A and B.
2. Only A and C.
3. Only C.
4. Only A.
26
MAT2611/101/3/2017
ASSIGNMENT 02
Due date: Thursday, 7 September 2017
Total Marks: 40
ONLY FOR SEMESTER 2
Answer all the questions. Show all your workings.
If you choose to submit via my Unisa, note that only PDF files will be accepted.
Let
1 1 0 1 0 −1 1 1 1 0 0 1
B1 = , , and B2 = , ,
1 −1 1 0 1 0 0 −1 1 −1 0 0
be two bases for span(B1 ) in M22 , where the usual left to right ordering is assumed.
(1.1) Find the transition matrix (change of coordinate/change of basis matrix) PB1 →B2 . (8)
(1.2) Let B3 be a basis for span(B1 ) and let the transition matrix from B2 to B3 be given by
 
1 1 0
PB2 →B3 = 1 0 1 .
1 1 1
(a) Find the transition matrix PB1 →B3 . (6)

(b) Use PB2 →B3 to find B3 . (6)

1 0 1 0
0 1 0 0
A=
0
.
1 0 0
1 0 0 1
(2.1) Determine the characteristic equation for A in λ. (4)
(2.2) Find the eigenvalues of A, and their algebraic multiplicities. (4)
(2.3) Find a basis for the eigenspace corresponding to each eigenvalue of A and hence also (12)
27
ASSIGNMENT 03
Due date: Thursday, 5 October 2017
ONLY FOR SEMESTER 2
Question 1
2. λ = 0.
4. A is invertible.
Question 2
Let A be an n × n matrix with eigenvalue −1, In be the n × n identity matrix and 0n be the n × n
zero matrix. Which of the following are true?
A. 0 is an eigenvalue of A + In .
B. A + In is singular.
C. A + In = 0n .
D. 1 is an eigenvalue of A2 .
1. Only A, B and D.
2. Only A, B and C.
3. Only A, C and D.
28
MAT2611/101/3/2017
Question 3

   
1 1 1 1 0 1
1 1 0 1
A. 0 2 2. B. 0 0 0. C. . D. .
1 0 1 1
0 0 3 0 0 1
1. Only A, C and D.
2. Only A.
3. Only A and B.
4. Only C and D.
Question 4
Let A and B be n × n matrices and let In be the n × n identity matrix. Then
1. If ABB T AT is diagonalizable.
2. If A is diagonalizable then A is invertible.
3. If λ = 0 is an eigenvalue of A, then A is not diagonalizable.
4. If A and B are not diagonalizable then A + B is not diagonalizable.
Question 5
1. hp(x), q(x)i = p(1)q(1) + 2p(2)q(2) + 3p(3)q(3) in P2 .

1 1 T
1 1
3. h(x1 , x2 ), (y1 , y2 )i = x1 y2 + x2 y1 in R2 .
4. h(x1 , x2 ), (y1 , y2 )i = x1 y1 + x2 y2 − 1 in R2 .
29
Question 6
h(x1 , x2 , x3 ), (y1 , y2 .y3 )i = 2x1 y1 + 2x2 y2 + x3 y3 in R3 ?
√ √
A. (1, 0, 0) B. (0, 1, 0)/ 2 C. (1, 1, 1)/ 3 D. (1, 1, 0)/2
1. Only B and D.
2. Only A, C and D.
3. Only A and C.
4. Only A.
Question 7
Which of thefollowing
vectors
are orthogonal to each other with respect to the inner product
1 0 T
hA, Bi = tr A B in M22 ?
0 2

1 1 1 1 1 −1 1 1
A. . B. . C. . D. .
1 1 1 −1 −1 1 −1 −1
1. Only A and C are orthogonal, A and D are orthogonal, C and D are orthogonal.
2. Only B and C are orthogonal.
3. Only A and C are orthogonal, B and D are orthogonal.
4. Only A and C are orthogonal, A and D are orthogonal.
Question 8
Consider the vector subspace W = span{ 1 − x, 2x2 } of P2 with the standard inner product. Which
of the following vectors in P2 lie in the subspace W ⊥ of P2 ?
1. x2 + 1.
2. x + 1.
3. x − 1.
4. x2 − 1.
30
MAT2611/101/3/2017
ADDENDUM C: EXAM INFORMATION SHEET

The question papers include an information sheet. Please see myUnisa for past papers and their
information sheets. An example of an information sheet is reproduced below. The information sheet
includes all of the essential concepts and theorems.
INFORMATION SHEET
Vector spaces
Definition (Vector space).
A vector space is a non-empty set V with vector addition + : V × V → V and scalar multiplication
· : R × V → V obeying the axioms
VS1. u + v ∈ V for all u, v ∈ V ,
VS2. u + v = v + u for all u, v ∈ V ,
VS3. u + (v + w) = (u + v) + w for all u, v, w ∈ V ,
VS4. there exists 0 ∈ V such that u + 0 = u for all u ∈ V ,
VS5. for all u ∈ V there exists −u ∈ V such that u + (−u) = 0,
VS6. a · u ∈ V for all a ∈ R, u ∈ V ,
VS7. a · (u + v) = a · u + a · v for all a ∈ R, u, v ∈ V ,
VS8. (a + b) · u = a · u + b · u for all a, b ∈ R, u ∈ V ,
VS9. a · (b · u) = (ab) · u for all a, b ∈ R, u ∈ V ,
VS10. 1 · u = u for all u ∈ V .
Theorem (VZ). 0 = 0 · u = a · 0 for all a ∈ R and u ∈ V in a vector space V .
Theorem (VN). (−1) · u = −u for all u ∈ V in a vector space V .
Definition (Subspace).
A subset W ⊆ V of a vector space V is a subspace of V if W , with the same vector addition and
scalar multiplication as V , is a vector space.
Theorem (SS).
A subset W ⊆ V of a vector space V is a subspace of V , with the same vector addition + and scalar
multiplication · as V , if and only if
1. W is not empty,
2. u + v ∈ W for all u, v ∈ W ,
31
3. a · u ∈ W for all a ∈ R, u ∈ V .
Definition (Linear independence).

A subset {b1 , . . . , bn } ⊆ V in a vector space V is linearly independent if and only if
c1 · b1 + · · · + cn · bn = 0 ⇐⇒ c1 = · · · = cn = 0.
Definition (Span).
The span of a subset {b1 , . . . , bn } ⊆ V in a vector space V is the subspace of V given by
span{b1 , . . . , bn } = {c1 · b1 + · · · + cn · bn : c1 , . . . , cn ∈ R}.
Definition (Basis, dimension).

A subset {b1 , . . . , bn } ⊆ V in a vector space V is a basis for V if and only if
1. {b1 , . . . , bn } is linearly independent,
2. span{b1 , . . . , bn } = V .
If {b1 , . . . , bn } ⊆ V is a basis for V then the dimension of V is n, dim(V ) = n.
Definition (Coordinate matrix).

Let B = {b1 , . . . , bn } be a basis for V and let v ∈ V . Then there exists unique c1 , . . . , cn ∈ R such
that v = c1 · b1 + · · · + cn · bn . The column vector
 
c1
 .. 
[v]B =  . 
cn
is the coordinate matrix of v relative to B.
Definition (Transition matrix, change of coordinate matrix).

Let B1 = {b1 , . . . , bn } be a basis for the vector space V , and B2 be another basis for V . The transition
matrix (change of coordinate matrix) PB1 →B2 from B1 to B2 is given by

PB1 →B2 = [b1 ]B2 · · · [bn ]B2 .
Examples (of vector spaces).
• Rn
• The vector space Pn = {c0 + c1 x + · · · + cn xn : c0 , . . . , cn ∈ R} of polynomials of degree n or less.
• The vector space Mmn of m × n matrices.
32
MAT2611/101/3/2017
Inner products
Definition (Inner product).
An inner product is a function h·, ·i : V × V → R on a vector space V which obeys the axioms
IP1. hu, vi = hv, ui for all u, v ∈ V ,
IP2. hk · u, vi = khu, vi for all k ∈ R, u, v ∈ V ,
IP3. hu, v + wi = hu, vi + hu, wi for all u, v, w ∈ V ,
IP4. a) hu, ui ≥ 0, for all u ∈ V ,

b) hu, ui = 0 if and only if u = 0.
Definition (Orthogonality).
Let h·, ·i denote an inner product on a vector space V . If hu, vi = 0, then u and v are orthogonal to
each other.
Definition (Unit vector, normalized).

Let h·, ·i denote an inner product on a vector space V . If hu, ui = 1, then u is a unit vector (normal-
ized).
Theorem (Cauchy-Schwarz inequality).

Let h·, ·i denote an inner product on a vector space V . Then
p
|hu, vi| ≤ hu, uihv, vi
for all u, v ∈ V .
Definition (Gram-Schmidt process).

Let h·, ·i denote an inner product on a vector space V and let {u1 , . . . , um } be a linearly independent
set in V . The Gram-Schmidt process yields an orthogonal basis {v1 , . . . , vm } for span{u1 , . . . , um }
as follows
v1 = u1 ,
hu2 , v1 i
v2 = u2 − v1 ,
hv1 , v1 i
..
.
hum , v1 i hum , vm−1 i
vm = um − v1 − · · · − vm−1 .
hv1 , v1 i hvm−1 , vm−1 i
vj
An orthonormal basis {v10 , . . . , vm
0
} is obtained by setting vj0 = .
hvj , vj i
33
Linear transformations
Definition (Linear transformation).
A function T : V → W between vector spaces V and W is a linear transformation if and only if
1. T (k · u) = k · T (u) for all k ∈ R, u ∈ V
2. T (u + v) = T (u) + T (v) for all u, v ∈ V
Examples (of linear transformations).

• The trace operation on Mnn is a linear transformation tr : Mnn → R.
• The transpose operation on Mmn is a linear transformation.
Definition (Kernel, nullity).

The kernel of a linear transformation T : V → W between vector spaces V and W is the subspace
ker(T ) = {v ∈ V : T (v) = 0W }
of V , where 0W is the zero vector in W . The nullity of T is the dimension of ker(T ).
Definition (Range, rank).

The range of a linear transformation T : V → W between vector spaces V and W is the subspace
R(T ) = {T (v) : v ∈ V }
of W . The rank of T is the dimension of R(T ).
Definition (One-to-one, injective, inverse).

A linear transformation T : V → W between vector spaces V and W is one-to-one if and only if
T (u) = T (v) ⇐⇒ u = v.
A one-to-one linear transformation T : V → W has an inverse linear transformation T −1 : R(T ) → V

satisfying T −1 (T (u)) = u for all u ∈ V .
Definition (Onto, surjective).

A linear transformation T : V → W between vector spaces V and W is onto if and only if R(T ) = W .
Theorem (TO). If V and W are finite dimensional vector spaces and T : V → W is a linear
transformation, then T is one-to-one if and only if ker(T ) = {0}. If dim(V ) = dim(W ), then T is
onto if and only if T is one-to-one.
Definition (Isomorphism, bijection).

A one-to-one and onto linear transformation T : V → W between vector spaces V and W is an
isomorphism (bijection). If an isomorphism between V and W exists, then V and W are isomorphic.
34
MAT2611/101/3/2017
Theorem (VI). Every vector space V with dim(V ) = n is isomorphic to Rn .
Definition (Matrix representation of a linear transformation).

Let BV = {b1 , . . . , bn } be a basis for the vector space V , and BW be a basis for the vector space W .
The matrix representation [T ]BW ,BV of a linear transformation T : V → W is given by

[T ]BW ,BV = [T (b1 )]BW · · · [T (bn )]BW .
When V = W and BV = BW , we write [T ]BV = [T ]BV ,BV .
Matrices
Definition (Column space, row space, rank).  
r1
Let A be an m × n matrix with columns A = c1 · · · cn and rows A =  ... .
 
rm
The column space of A is span{c1 , . . . , cn } and the row space of A is span{r1 , . . . , rm }. The rank of A
is the dimension of the column and row spaces, rank(A) = dim(span{c1 , . . . , cn }) = dim(span{r1 , . . . , rm }).
Definition (Null space, nullity).

The null space of an m × n matrix A is the subspace
N (A) = {x ∈ Rn : Ax = 0}.
The nullity of T is the dimension of N (A).
Theorem (RN). rank(A) + nullity(A) = n for every m × n matrix A.
Definition (Eigenvalue, eigenvector).

Let A be an n × n matrix. If Ax = λx, for λ ∈ C and x ∈ Cn with x 6= 0, then λ is an eigenvalue of
A and x is an eigenvector of A corresponding to the eigenvalue λ.
Definition (Eigenspace, geometric multiplicity).

Let A be an n × n matrix, and let λ be an eigenvalue of A. Then
Eλ = {x ∈ Cn : Ax = λx}
is a vector space, called the eigenspace for the eigenvalue λ of A. The geometric multiplicity of λ is
dim(Eλ ).
Definition (Characteristic equation, characteristic polynomial).

Let A be an n × n matrix. Then λ ∈ C is an eigenvalue of A if and only if λ satisfies the characteristic
equation det(λIn − A) = 0, where In is the n × n identity matrix. The polynomial det(xIn − A) is the
characteristic polynomial in the variable x.
35
Definition (Algebraic multiplicity).
Let A be an n × n matrix with eigenvalue λ. The algebraic multiplicity of λ is the largest number
a ∈ N such that (x − λ)a is a factor of the characteristic polynomial det(xIn − A).
Definition (Diagonalizable).
An n × n matrix A is diagonalizable if and only if A is similar to some n × n diagonal matrix D, i.e.
A = P DP −1 for some n × n diagonal matrix D and non-singular n × n matrix P .
Theorem (DI). An n × n matrix A is diagonalizable if and only if A has n linearly independent

eigenvectors.
Theorem (DD). If an n × n matrix A has n distinct eigenvalues, then A is diagonalizable.
Theorem (DS). If an n × n matrix A is symmetric, then A is diagonalizable.
Theorem (DM). For a square matrix A, the algebraic and geometric multiplicity are equal for each
eigenvalue of A if and only if A is diagonalizable.
Definition (Trace).
The trace of a square matrix is the sum of its diagonal entries
 
a11 a12 · · · a1n
 a21 a22 · · · a2n 
tr  .. ..  = a11 + a22 + · · · + ann .
 
.. . .
 . . . . 
an1 an2 · · · ann
Theorem (CT). For all n × n matrices A, B and C we have tr(ABC) = tr(CAB). Consequently
tr(AB) = tr(BA).
Definition (Transpose).
The transpose of a matrix is obtained by interchanging corresponding rows and columns
 T  
a11 a12 · · · a1n a11 a21 · · · am1
 a21 a22 · · · a2n   a12 a22 · · · am2 
= ..  .
   
 .. .. . . .
.   .. .. . .
 . . . .   . . . . 
am1 am2 · · · amn a1n a2n · · · amn
Theorem (TT). For all m × n matrices A we have (AT )T = A.
Theorem (TI). For all n × n matrices A we have tr(A) = tr(AT ).
36
MAT2611/101/3/2017
Determinants
For 2 × 2 and 3 × 3 matrices:
a b c
a b a b
det = = ad − bc, d e f = aei + bf g + cdh − af h − bdi − ceg
c d c d
g h i
Cofactor expansion along the j-th row:
a11 a12 ··· a1,k−1 a1,k+1 ··· a1n

a21 a22 ··· a2,k−1 a2,k+1 ··· a2n
a11 a12 · · · a1n .. .. ... .. .. ... ..
a21 a22 · · · a2n Xn . . . . .
.. .. .. .. = (−1)(j−1)+(k−1) ajk aj−1,1 aj−1,2 · · · aj−1,k−1 aj−1,k+1 · · · aj−1,n
. . . . k=1 aj+1,1 aj+1,2 · · · aj+1,k−1 aj+1,k+1 · · · aj+1,n
an1 an2 · · · ann .. .. .. .. .. .. ..
. . . . . . .
an1 an2 · · · an,k−1 an,k+1 · · · ann
Cofactor expansion along the k-th column:
a11 a12 ··· a1,k−1 a1,k+1 ··· a1n

a21 a22 ··· a2,k−1 a2,k+1 ··· a2n
a11 a12 · · · a1n .. .. ... .. .. ... ..
a21 a22 · · · a2n Xn . . . . .
.. .. ... .. = (−1)(j−1)+(k−1) ajk aj−1,1 aj−1,2 · · · aj−1,k−1 aj−1,k+1 · · · aj−1,n
. . . j=1 aj+1,1 aj+1,2 · · · aj+1,k−1 aj+1,k+1 · · · aj+1,n
an1 an2 · · · ann .. .. .. .. .. .. ..
. . . . . . .
an1 an2 · · · an,k−1 an,k+1 · · · ann
Theorem (DC). For all k ∈ R and n × n matrices A we have det(kA) = k n det(A).
Theorem (DP). For all n × n matrices A and B we have det(AB) = det(A) det(B).
37
ADDENDUM D: USEFUL COMPUTER SOFTWARE
It is possible to check the correctness of your calculations by hand. If you are interested in software
that may help to check your results please consult the following resources. Note however that the
software will not be available at exam time, so it is recommended to be proficient at
checking your own results by hand.
Maxima:
http://maxima.sourceforge.net/
http://maxima.sourceforge.net/docs/intromax/intromax.html (section 6).
http://maxima.sourceforge.net/docs/manual/en/maxima_23.html
Maxima is also available for Android devices:

https://sites.google.com/site/maximaonandroid/
See addendum E for a brief introduction to Maxima for Linear Algebra.
Wolfram Alpha:
http://www.wolframalpha.com/
http://www.wolframalpha.com/examples/Matrices.html
Please note that the use of software is not required for this module.
38
MAT2611/101/3/2017
ADDENDUM E: ELEMENTARY LINEAR ALGEBRA USING MAXIMA

A complete guide to Maxima is beyond the scope of this module. Here we list only the most essential
features. Please consult http://maxima.sourceforge.net/ for documentation on Maxima.
Please note that the use of software is not required for this module.
E.1 The linearalgebra and eigen packages

First we load the packages eigen and linearalgebra. Type only the line following (%i1) in the
white boxes, i.e. load(eigen);
(%i1) load(eigen);
(%o1) /usr/pkg/share/maxima/5.27.0/share/matrix/eigen.mac
(%i2) load(linearalgebra);
0 errors, 0 warnings
(%o2) /usr/pkg/share/maxima/5.27.0/share/linearalgebra/linearalgebra.mac
The output (%o1) and (%o2) and may be different, but there should be no error messages. Note the
semicolon ; after every command.
E.2 Matrices
Now we input the matrices  
−1 −2
1 2 3
A= , B= 1 2 .
4 5 6
0 0
(%i3) A: matrix( [1, 2, 3],
[4, 5, 6] );
[ 1 2 3 ]
(%o3) [ ]
[ 4 5 6 ]
(%i4) B: matrix( [-1, -2],

[ 1, 2],
[ 0, 0] );
[ - 1 - 2 ]
[ ]
(%o4) [ 1 2 ]
[ ]
[ 0 0 ]
39
Type carefully to reproduce the input (%i3) and (%i4) correctly. Next we calculate the matrix
product C = AB. The matrix product is denoted by a full stop between A and B.
(%i5) C: A . B;
[ 1 2 ]
(%o5) [ ]
[ 1 2 ]
E.3 Eigenvalues and eigenvectors

We can determine the eigenvalues of C, namely 0 and 3 each with algebraic multiplicity 1. The
expression eigenvalues(C) returns a list of eigenvalues [0, 3] and a list of multiplicities for each
eigenvalue [1, 1] where the multiplicities are in the same order as the eigenvalues.
(%i6) eigenvalues(C);
(%o6) [[0, 3], [1, 1]]
Similarly the eigenvectors eigenvectors(C) can be determined. This returns three lists, the first two
are the same as for eigenvalues(C) while the last is a list of eigenvectors.
(%i7) eigenvectors(C);
1
(%o7) [[[0, 3], [1, 1]], [[[1, - -]], [[1, 1]]]]
2
i.e. we find the eigenvector

1
− 12
for the corresponding eigenvalue 0 of C and the eigenvector

1
1
for the corresponding eigenvalue 3 of C. The normalized eigenvectors (uniteigenvectors(C)) can
be determined similarly.
(%i8) uniteigenvectors(C);
2 1 1 1
(%o8) [[[0, 3], [1, 1]], [[[-------, - -------]], [[-------, -------]]]]
sqrt(5) sqrt(5) sqrt(2) sqrt(2)
i.e. the normalized eigenvector corresponding to the eigenvalue 0 of C is

" #
√2
5
.
− √15
Although you may find a different eigenvector, that does not mean your answer is incorrect!
40
MAT2611/101/3/2017
E.4 Rank, nullity, columnspace and nullspace

The rank of A (appears above) is calculated with rank(A), the nullity with nullity(A), the columnspace
with columnspace(A) and the nullspace with nullspace(A). Once again, your own answers may differ
but still be correct!
(%i9) rank(A);
(%o9) 2
(%i10) columnspace(A);
[ 1 ] [ 2 ]
(%o10) span([ ], [ ])
[ 4 ] [ 5 ]
(%i11) nullspace(A);
[ - 3 ]
[ ]
(%o11) span([ 6 ])
[ ]
[ - 3 ]
(%i12) nullity(A);
(%o12) 1
E.5 Matrix inverse

The inverse of a matrix (when it exists) is calculated using invert. Here we calculate
−1
1 1
.
1 2
(%i13) invert(matrix( [1,1], [1,2] ));
[ 2 - 1 ]
(%o13) [ ]
[ - 1 1 ]
E.6 Gram-Schmidt algorithm

The Gram-Schmidt algorithm is easily applied using gramschmidt. The vectors for which we want
to find an orthogonal basis are specified as rows of a matrix. For example, below we apply the
gram-Schmidt algorithm for
41

1 0
u1 = , u2 =
1 1
with respect to the Euclidean inner product.
(%i14) gramschmidt(matrix([1,1],[0,1]));
1 1
(%o14) [[1, 1], [- -, -]]
2 2
i.e. we find the orthogonal basis 1

1 −
, 12 .
1 2
Now consider a non-Euclidean inner product on R2
hx, yi := x1 y1 + 2x2 y2 , x = (x1 , x2 ), y = (y1 , y2 ), x1 , x2 , y1 , y2 ∈ R
(%i15) f(x,y):= x[1]*y[1] + 2*x[2]*y[2];
(%o15) f(x, y) := x y + 2 x y
1 1 2 2
we can tell gramschmidt to use f (our inner product) when applying the Gram-Schmidt algorithm
(%i16) ob: gramschmidt(matrix([1,1],[0,1]), f);
2 1
(%o16) [[1, 1], [- -, -]]
3 3
i.e. we find the orthogonal basis 2

1 −
, 13 .
1 3
with respect to our non-Euclidean inner product. To find an orthonormal basis we need to normalize
each of these vectors with respect to our non-Euclidean inner product by extracting each vector and
divide by its norm. Here we use first, second and so on to obtain each of the vectors.
(%i17) v1: first(ob);
(%o17) [1, 1]
(%i18) v1 / sqrt(f(v1,v1));
42
MAT2611/101/3/2017
1 1
(%o18) [-------, -------]
sqrt(3) sqrt(3)
(%i19) v2: second(ob);
2 1
(%o19) [- -, -]
3 3
(%i20) v2 / sqrt(f(v2,v2));
2 1
(%o20) [- -------------, -------------]
4 2 4 2
3 sqrt(- + -) 3 sqrt(- + -)
9 9 9 9
To simplify the rational expressions, use ratsimp.

(%i21) ratsimp(v2 / sqrt(f(v2,v2)));
sqrt(2) 1
(%o21) [- -------, ---------------]
sqrt(3) sqrt(2) sqrt(3)
43
ADDENDUM F: Example questions
The following sections provide complete solutions for example past assignment and exam questions.
It is highly recommended to attempt the problems on your own before consulting these solutions. If
your own solutions are not consistent with those provided, attempt to identify where (and why) things
went awry. Note that some questions have many different, correct, answers.
F.1 Previous assignment questions

Questions
Question 1
Show that the set X with the given operations fails to be a vector space by identifying all axioms that
hold and fail to hold:
The set X = R3 with vector addition ⊕ defined by
(a, b, c) ⊕ (x, y, z) = (1, y, c + z)
and scalar multiplication defined by k (a, b, c) = (ka, kb, kc).
Question 2
The set X = {(a, 1) : a ∈ R} ⊂ R2 with vector addition ⊕ defined by
(a, 1) ⊕ (b, 1) = (a + b, 1)
and scalar multiplication defined by k (a, 1) = (k 2 a, 1).
Question 3
Consider the set

X := { (1, x) : x ∈ R }
and the operations (for all k, x, y ∈ R, a = (1, x) ∈ X and b = (1, y) ∈ X)
· : R × X → X, k · a ≡ k · (1, x) := (1, kx),

+ : X × X → X, a + b ≡ (1, x) + (1, y) := (1, x + y).
The set X with these definitions of · and + forms a vector space.
(3.1) Find the zero vector for X.
44
MAT2611/101/3/2017
(3.2) Prove that each axiom for vector spaces holds for X with the given operations.
(3.3) Is the vector space X a subspace of R2 ? Motivate your answer.
Question 4
Consider the set

X := { ex : x ∈ R }
and the operations (for all k, x, y ∈ R, a = ex ∈ X and b = ey ∈ X)
: R × X → X, k a ≡ k ex := (ex )k = ekx ,
⊕ : X × X → X, a ⊕ b ≡ ex ⊕ ey := ex · ey = ex+y .
The set X with these definitions of and ⊕ forms a vector space. Here we use instead of · and ⊕
instead of + to avoid confusion with the usual arithmetic operations.
(4.1) Find the zero vector for X. Find the zero vector for X.
(4.3) Is the vector space X a subspace of R ? Motivate your answer.
Question 5
Show that
Y := { (0, y) : y ∈ R },
with the usual vector addition and scalar multiplication in R2 , is a subspace of R2 .
Question 6
Let B ∈ M22 be a fixed but arbitrary 2 × 2 matrix. Show that
Y := { A : A ∈ M22 , AB = BA },
with the usual vector addition and scalar multiplication in M22 , is a subspace of M22 .
Question 7
Is
1 2 1 1 2
span , , = span , ?
2 1 1 2 1
45
Question 8
Is
1 2 1
span , ,
2 1 1
a linearly independent subset of R2 ? Motivate your answer.
Question 9
Consider the vector space M22 of all 2 × 2 matrices.
(9.1) Show that B = {A1 , A2 , A3 , A4 } is a basis for M22 where

3 6 0 −1 0 −8 1 0
A1 = ; A2 = ; A3 = ; and A4 = .
3 −6 −1 0 −12 −4 −1 2

6 2
(9.2) Write A = as a linear combination of vectors from B.
5 3
(9.3) Prove that the subset

D = {A ∈ M22 : AT + A = 0}
forms a subspace of M22 .
Question 10

3 −2 1 −1 0 1 1 0
A1 = ; A2 = ; A3 = ; and A4 = .
0 1 0 0 1 0 0 0

6 2
5 3

D = {A ∈ M22 : tr(A) = 0}
Question 11
Given the matrix

1 2 1
A=
2 1 1
(11.1) Find the range of A.
46
MAT2611/101/3/2017
(11.2) Find the rank of A.
(11.3) Find the null space of A.
(11.4) Find the nullity of A.
Question 12
Given the matrix  

1 2
A = 2 1
1 1
(12.3) Find the nullspace of A.
Question 13
(13.1) Let S and T be subspaces of a vector space V . Prove that S ∩ T is also a subspace of
V.
(13.2) Let S = {(a + b, a, −a, −b) : a, b ∈ R} and T = {(x, 2y, −y, −x) : x, y ∈ R}.
(a) Show that S and T are subspaces of R4 .
(b) Find the dimension of S and T .
(c) Find S ∩ T and hence a basis and the dimension of S ∩ T .
(13.3) Let
S = {1 − x; 5 + 3x − 2x2 ; 1 + 3x − x2 } ⊂ P2 .
Find a basis and the dimension for span(S).
Question 14
(14.1) Let S = {(a, b, b, a + c) : a, b, c ∈ R} and T = {(x, y − x, y + z, z) : x, y, z ∈ R}.

47
(14.2) Let
S = {t3 + t2 − 2t + 1; t2 + 1; t3 − 2t; 2t3 + 3t2 − 4t + 3} ⊂ P3 .
Question 15
Let B = {p1 , p2 } and B 0 = {q1 , q2 } be basis for P1 , where
p1 = 1 + 2x, p2 = 3 − x, q1 = 2 − 2x, q2 = 4 + 3x.
(15.1) Find the transition matrix PB 0 −→B from the B 0 -basis to the B-basis
(15.2) Find the transition matrix QB 0 −→B from the B-basis to the B 0 -basis
(15.3) Compute [p]B 0 if p = 5 − x.
Question 16
p1 = 1 + 3x, p2 = 2 − x, q1 = 1 − 2x, q2 = 1 + x.
(16.3) Compute [p]B 0 if p = 3 + x.
Question 17
Let B1 = { 1 + x, 1 − x } ⊂ P1 and B2 = { 1 + 2x, 2 + x } ⊂ P1 be two bases for P1 , where the usual

left to right ordering is assumed.
(17.1) Show that B2 is a basis for P1 .
(17.2) Find the transition matrix PB1 →B2 .
(17.3) Let B3 be a basis

for P1 and PB2 →B3 be the transition matrix from B2 to B3 given by
1 1
PB2 →B3 = .
0 1
(a) Find the transition matrix PB1 →B3 .
(b) Use PB2 →B3 to find B3 .
48
MAT2611/101/3/2017
Question 18


1 1
PB1 →B3 = .
0 2

Question 19

1 2 3
A = 1 2 3 .
1 2 3
(19.1) Determine the characteristic equation for A in λ.
(19.2) Find the eigenvalues of A, and their algebraic multiplicities.
(19.3) Find a basis for the eigenspace corresponding to each eigenvalue of A and hence also
Question 20
 
Consider the matrix 1 2 0
A = 2 1 0  .
0 0 −1
49
Question 21
Let  
−1 0 1
A= 0 2 0
0 −3 1
(21.1) Show that A is diagonalizable.
(21.2) Find an invertible matrix P and a diagonal matrix D such that P −1 AP = D.
(21.3) Calculate A11 .
Question 22
Let  
1 5 0
A = 5 1 0
1 −1 6
Question 23
Consider the matrix (see 19)  

1 2 3
A = 1 2 3 .
1 2 3
(23.1) Find an invertible matrix P such that D := P −1 AP is diagonal. Determine D.
(23.2) Find the rank of D and hence also the rank of A.
(23.3) Calculate Dn for n ∈ N and hence also An as a matrix.
(23.4) Show that  

1
A = 1 1 2 3 .
1
Use this expression for A to calculate A2 , A3 etc. and compare with your answer to
(24.3).
50
MAT2611/101/3/2017
Question 24

1 2 0
A = 2 1 0  .
0 0 −1
(24.1) Find an orthogonal matrix P such that D := P T AP is diagonal. Determine D.
(24.4) Let B be a m × m matrix where m ∈ N, I be the m × m identity matrix and k ∈ R.

(a) Let x be an eigenvector of B with corresponding eigenvalue λ. Show that x is an
eigenvector of B + kI. What is the corresponding eigenvalue of B + kI?
(b) Assume that B is diagonalizable, is B + kI diagonalizable?
Question 25
Consider the vector space R3 .
(25.1) Show that

   
x1 y1
hx, yi := 3x1 y1 + x2 y2 + x3 y3 , x = x2 , y = y2  ∈ R3
  
x3 y3
is an inner product on R3 .
(25.2) Apply the Gram-Schmidt process to the following subset of R3 :

     
 1 −1 −1 
0 ,  0  ,  1 
1 1 1
 
to find an orthonormal basis with respect to the inner product defined in 2.1 for the
span of this subset.
Question 26
(26.1) Show that

   
x1 y1
hx, yi := x1 y1 + 3x2 y2 + x3 y3 , x = x2 , y = y2  ∈ R3
  
x3 y3
51
     
 1 −1 −1 
0 ,  0  ,  1 
1 1 1
 
Question 27
(27.1) Show that

hu, vi = u1 v1 + 2u2 v2 + 3u3 v3
is an inner product on R3 for u = (u1 , u2 , u3 ) and v = (v1 , v2 , v3 ).
(27.2) Let u = (1, 1, 1), v = (1, 1, 0) and w = (1, 0, 0). Show that B = {u, v, w} is linearly
independent and spans R3 .
(27.3) Transform B into an orthonormal basis using the inner product in 27.1.
(27.4) Let R3 have the Euclidean inner product and W = span{u, v} where

4 3
u= , 0, − and v = (0, 1, 0).
5 5
Express w = (1, 2, 3) in the form w = w1 + w2 where w1 ∈ W and w2 ∈ W ⊥ .
Question 28
(28.1) Show that

hu, vi = 2u1 v1 + 3u2 v2 + u3 v3
(28.2) Let u = (1, 1, 1), v = (−1, 1, 0) and w = (1, 2, 1). Show that B = {u, v, w} is linearly
u = (1, 0, −1) and v = (3, 1, 0).
52
MAT2611/101/3/2017
Question 29
(29.1) Let
 
0 0 2 0
1 0 1 0
A= .
0 1 −2 0
0 0 0 1
Find the bases for the eigenspaces associated with the eigenvalues of A.
(29.2) Let T : P2 → P2 be the function defined by T (p(x)) = p(2x + 1).

(a) Show that T is a linear transformation.
(b) Find [T ]B with respect to the basis {1, x, x2 }.
(c) Compute T (2 − 3x + 4x2 ).
(29.3) Let S : P2 → P3 be defined by S(p(x)) = xp(x).

(a) Show that S is one-to-one.
(b) Find S −1 (p(x)).
(c) Is S onto? Explain.
Question 30
(30.1) Let
 
−2 0 0 0
 0 −2 5 −5 
A=
0
.
0 3 0 
0 0 0 3
(30.2) Let T : P2 → P3 be the function defined by T (p(x)) = xp(x − 3).

(b) Find [T ]B 0 ,B with respect to the basis B = {1, x, x2 } and B 0 = {1, x, x2 , x3 }.
(c) Compute T (1 + x − x2 ).
(30.3) Let S : P3 → P3 be defined by S(p(x)) = p(x + 1).

(b) Find S −1 (p(x)).
53
Question 31
Let T : P2 → P2 be defined by
T (a0 + a1 x + a2 x2 ) = (2a0 − a1 + 3a2 ) + (4a0 − 5a1 )x + (a1 + 2a2 )x2 .
(31.1) Find the eigenvalues of T .
(31.2) Find bases for the eigenspaces of T .
Question 32
T (a0 + a1 x + a2 x2 ) = −2a2 + (a0 + 2a1 + a2 )x + (a0 + 3a2 )x2 .
Question 33
Let T : R3 → R3 be the linear operator given by

   
x −x + 2y + 4z
T y  =  3x + z .
z 2x + 2y + 5z
(33.1) Find a basis B 0 for R3 relative to which the matrix T is diagonal using the standard
basis B for R3 .
(33.2) Compute [T ]B 0 and verify that [T ]B 0 = P −1 [T ]B P where the matrix P diagonalizes

[T ]B .
Question 34

   
x 4x + z
T y  = 2x + 3y + 2z  .
z x + 4z
basis B for R3 .
54
MAT2611/101/3/2017
(34.2) Compute [T ]B 0 and verify that [T ]B 0 = P −1 [T ]B P where the matrix P diagonalizes

[T ]B .
Question 35
 
−1 2 4
Let T : R3 → R3 be multiplication by the matrix A =  3 0 1 . Find
2 2 5
(35.1) a basis for the range of T ,
(35.2) a basis for the kernel of T ,
(35.3) the rank and nullity of T and
(35.4) the rank and nullity of A.
Question 36
 
1 0 3
Let T : R3 → R3 be multiplication by the matrix A = 1 2 4 . Find
1 8 5
Question 37

1 2a b
Consider T : P2 → M22 given by T (a + bx + cx2 ) = for all a, b, c ∈ R.
2 b 2c
(37.1) Show that T is a linear transform.
(37.2) Find the matrix representation for T relative to the standard basis in P2 and in M22
with the usual ordering.
(37.3) Is T invertible?
(37.4) Show that the range of T is the subspace M

f22 of M22 consisting of symmetric matrices.
55
(37.5) Let Te : P2 → M
f22 be defined by Te(p(x)) := T (p(x)) for all p(x) ∈ P2 . Find the matrix
representation for Te relative to the standard basis with the usual ordering in P2 and the
basis
1 0 0 0 0 1
, ,
0 0 0 1 1 0
for the 2 × 2 symmetric matrices, ordered left to right.
Question 38
Consider T : M22 → P2 given by

1
T (A) = 1 x A , for all A ∈ M22 .
x
(38.2) Find the matrix representation for T relative to the standard basis in M22 and in P2
(38.3) Is T one to one?
(38.4) Let Mf22 be the subspace of M22 consisting of symmetric matrices. Let Te : Mf22 → P2 be
defined by Te(A) := T (A) for all A ∈ M
f22 . Find the matrix representation for Te relative
to the standard basis with the usual ordering in P2 and the basis

1 0 0 0 0 1
, ,
0 0 0 1 1 0
Question 39
Consider T : P2 → P2 given by T (a + bx + cx2 ) = b + cx + ax2 for all a, b, c ∈ R.
(39.1)
(a) Find the kernel and nullity of T .
(b) Find the range and rank of T .
(39.2) Find the real valued eigenvalues and corresponding eigenspaces of T .
(39.3) Find T 3 := T ◦ T ◦ T .
56
MAT2611/101/3/2017
Solutions
Question 1
The set X = R3 with vector addition ⊕ defined by
(a, b, c) ⊕ (x, y, z) = (1, y, c + z)
and scalar multiplication defined by k (a, b, c) = (ka, kb, kc).

In the following let u = (u1 , u2 , u3 ) ∈ R3 , v = (v1 , v2 , v3 ) ∈ R3 , w = (w1 , w2 , w3 ) ∈ R3 be arbitrary
elements of R3 and k, m ∈ R . We have
1. u ⊕ v = (1, v2 , u3 + v3 ) ∈ R3 holds.
2. u ⊕ v = (1, v2 , u3 + v3 ).
v ⊕ u = (1, u2 , v3 + u3 ).
Choosing u = (1, 0, 0) and v = (0, 1, 0) we see that u ⊕ v = v ⊕ u does not hold in general.
3. u ⊕ (v ⊕ w) = u ⊕ (1, w2 , v3 + w3 ) = (1, w2 , u3 + v3 + w3 ).
(u ⊕ v) ⊕ w = (1, v2 , u3 + v3 ) ⊕ w = (1, w2 , u3 + v3 + w3 ).
Thus u ⊕ (v ⊕ w) = (u ⊕ v) ⊕ w holds.
4. Suppose the zero vector 0 = (a, b, c) ∈ R3 exists. Then

u ⊕ 0 = (u1 , u2 , u3 ) ⊕ (a, b, c) = (1, b, u3 + c) = (u1 , u2 , u3 ) ⇔ u1 = 1, u2 = b, c = 0.
The zero vector property obviously does not hold, in particular the equation cannot be satisfied
for u = (0, 0, 0).
5. Since the zero vector does not exist, the negative is undefined. Thus the existence of negatives
does not hold.
6. k u = (ku1 , ku2 , ku3 ) ∈ R3 holds.
7. k (u + v) = k (1, v2 , u3 + v3 ) = (k, kv2 , ku3 + kv3 ).

k u ⊕ k v = (1, kv2 , ku3 + kv3 ).
Choosing k = 2, for example, we find k (u + v) = k u + k v does not hold in general.
8. (k + m) u = ((k + m)u1 , (k + m)u2 , (k + m)u3 ).

k u ⊕ m u = (1, mu2 , ku3 + mu3 ).
Choosing k = m = 0, for example, we find (k + m) u=k u⊕m u does not hold in
general.
9. k(m u) = k (mu1 , mu2 , mu3 ) = (kmu1 , kmu2 , kmu3 ).

(km) u = ((km)u1 , (km)u2 , (km)u3 ).
Thus k(m u) = (km) u holds.
10. 1 u = (1 · u1 , 1 · u2 , 1 · u3 ) = u holds.
57
Question 2
The set X = {(a, 1) : a ∈ R} ⊂ R2 with vector addition ⊕ defined by

(a, 1) ⊕ (b, 1) = (a + b, 1)
and scalar multiplication defined by k (a, 1) = (k 2 a, 1).
In the following let u = (a, 1) ∈ X, v = (b, 1) ∈ X, w = (c, 1) ∈ X be arbitrary elements of X and
k, m ∈ R. We have
1. u ⊕ v = (a + b, 1) ∈ X holds.
2. u ⊕ v = (a + b, 1).
v ⊕ u = (b + a, 1) = (a + b, 1).
Thus u ⊕ v = v ⊕ u holds.
3. u ⊕ (v ⊕ w) = u ⊕ (b + c, 1) = (a + (b + c), 1) = (a + b + c, 1).
(u ⊕ v) ⊕ w = (a + b, 1) ⊕ w = ((a + b) + c, 1) = (a + b + c, 1).
Thus u ⊕ (v ⊕ w) = (u ⊕ v) ⊕ w holds.
4. Suppose the zero vector 0 = (z, 1) ∈ X exists. Then
u ⊕ 0 = (a, 1) ⊕ (z, 1) = (a + z, 1) = (a, 1) ⇔ a + z = a ⇔ z = 0.
The zero vector property holds with 0 = (0, 1).
5. If X is a vector space then
u + (−1) u = 1 u + (−1) u = (1 − 1) u = 0 u = 0,
i.e. −u ≡ (−1) u.
However, here we must find −u directly. Let −u := (b, 1) ∈ X, b ∈ R. Then
−u + u = (b, 1) + (a, 1) = (b + a, 1) = (0, 1) = 0.
Thus b = −a. From
u + (−u) = (a, 1) + (b, 1) = (a + b, 1) = (0, 1) = 0.
it also follows that b = −a. Thus −u = −(a, 1) = (−a, 1) exists.
6. k u = (k 2 a, 1) ∈ X holds.
7. k (u + v) = k (a + b, 1) = (k 2 (a + b), 1) = (k 2 a + k 2 b, 1).
k u + k v = (k 2 a, 1) + (k 2 b, 1) = (k 2 a + k 2 b, 1).
Thus k (u + v) = k u + k vholds.
8. (k + m) u = ((k + m)2 a, 1) = (k 2 a + 2kma + m2 a, 1).
k u ⊕ m u = (k 2 a, 1) ⊕ (m2 a, 1) = (k 2 a + m2 a, 1).
Choosing k = m = 1, for example, we find (k + m) u = k u⊕m u does not hold in
general.
9. k(m u) = k (m2 a, 1) = (k 2 (m2 )a, 1) = (k 2 m2 a, 1).
(km) u = ((km)2 a, 1) = (k 2 m2 a, 1).
Thus k(m u) = (km) u holds.
10. 1 u = (1 · a, 1) = (a, 1) = u holds.
58
MAT2611/101/3/2017
Question 3
Consider the set

X := { (1, x) : x ∈ R }
and the operations (for all k, x, y ∈ R, a = (1, x) ∈ X and b = (1, y) ∈ X)
· : R × X → X, k · a ≡ k · (1, x) := (1, kx),

+ : X × X → X, a + b ≡ (1, x) + (1, y) := (1, x + y).
The set X with these definitions of · and + forms a vector space.
(3.1) Find the zero vector for X.

Let 0 ∈ X be the zero vector, i.e. there exists z ∈ R such that 0 = (1, z) and for all
a = (1, x) ∈ X (x ∈ R)
0 + a = (1, z) + (1, x) = (1, z + x) = (1, x) = a,

a + 0 = (1, x) + (1, z) = (1, x + z) = (1, x) = a.
Notice that we only use the definition of + given in the question. Since these two
equations relate pairs of real numbers, we can equate each component, i.e. 1 = 1 (holds
trivially), z + x = x and x + z = x. By subtracting the real number x from both sides
of each equation we find that z = 0 (since z + x − x = x + z − x = z and x − x = 0
for x, z ∈ R). Thus 0 = (1, 0) for X with vector addition and scalar multiplication
as defined in the question. (Notice that the zero vector does not depend on x, and
therefore also not on a; otherwise it would not satisfy axiom 4 in Definition 1 on page
172 of the textbook.)
1. Let a = (1, x) ∈ X and b = (1, y) ∈ X, i.e. x, y ∈ R. Then a + b = (1, x + y) ∈ X

since the first value of the pair is 1, and the second x + y ∈ R is a real number.
2. Let a = (1, x) ∈ X and b = (1, y) ∈ X. Then
a + b = (1, x) + (1, y) = (1, x + y)

b + a = (1, y) + (1, x) = (1, y + x)
so that a + b = b + a (since the addition of real numbers is commutative, i.e.

x + y = y + x).
3. Let a = (1, x) ∈ X, b = (1, y) ∈ X and c = (1, z) ∈ X. Then
(a + b) + c = (1, x + y) + (1, z) = (1, (x + y) + z) = (1, x + y + z)

a + (b + c) = (1, x) + (1, y + z) = (1, x + (y + z)) = (1, x + y + z)
due to the associativity of real numbers, i.e. (x + y) + z = x + (y + z). Thus

(a + b) + c = a + (b + c).
59
4. The existence of the zero vector is demonstrated in Question (4.1).
5. Let a = (1, x) ∈ X. Assume there exists x0 ∈ R such that −a = (1, x0 ), i.e.
a + (−a) = (1, x) + (1, x0 ) = (1, x + x0 ) = (1, 0) = 0

(−a) + a = (1, x0 ) + (1, x) = (1, x0 + x) = (1, 0) = 0.
Since these two equations relate pairs of real numbers, we can equate each com-
ponent, i.e. 1 = 1 (holds trivially), x + x0 = 0 and x0 + x = 0. Thus x0 = −x.
Consequently −a exists and is given by −a = −(1, x) = (1, −x).
6. Let k ∈ R and a = (1, x) ∈ X (so that x ∈ R). Then k · a = (1, kx) ∈ X since the
first value of the pair is 1, and the second kx ∈ R is a real number.
7. Let k ∈ R, a = (1, x) ∈ X and b = (1, y) ∈ X. Then
k · (a + b) = k · ((1, x) + (1, y)) = k(1, x + y) = (1, k(x + y)) = (1, kx + ky)

(k · a) + (k · b) = (1, kx) + (1, ky) = (1, kx + ky)
so that k · (a + b) = k · a + k · b.
8. Let k, m ∈ R and a = (1, x) ∈ X. Then
(k + m) · a = (1, (k + m)x) = (1, kx + mx)

(k · a) + (m · a) = (1, kx) + (1, mx) = (1, kx + mx)
so that (k + m) · a = k · a + m · a.
9. Let k, m ∈ R and a = (1, x) ∈ X. Then
(km) · a = (1, (km)x) = (1, kmx)

k · (m · a) = k · (1, mx) = (1, k(mx)) = (1, kmx)
due to the associativity of products of real numbers. Thus (km) · a = k · (m · a).
10. Let a = (1, x) ∈ X. Then
1 · a = 1 · (1, x) = (1, 1x) = (1, x) = a.
(3.3) Is the vector space X a subspace of R2 ? Motivate your answer.

A subspace X of a vector space V is a subset of V together with the same vector addition
and scalar multiplication as in V . Here we use a vector addition for X which is different
from the usual vector addition in R2 . Hence X is not a subspace of R2 .
60
MAT2611/101/3/2017
Question 4
Consider the set

X := { ex : x ∈ R }
and the operations (for all k, x, y ∈ R, a = ex ∈ X and b = ey ∈ X)
: R × X → X, k a ≡ k ex := (ex )k = ekx ,
⊕ : X × X → X, a ⊕ b ≡ ex ⊕ ey := ex · ey = ex+y .
The set X with these definitions of and ⊕ forms a vector space. Here we use instead of · and ⊕
instead of + to avoid confusion with the usual arithmetic operations.
(4.1) Find the zero vector for X. Find the zero vector for X.
Let 0 ∈ X be the zero vector, i.e. there exists z ∈ R such that 0 = ez and for all
a = ex ∈ X (x ∈ R)
0 ⊕ a = ez ⊕ ex = ez+x = ex = a,
a ⊕ 0 = ex ⊕ ez = ex+z = ex = a.
Notice that we only use the definition of ⊕ given in the question. Since the exponential
function is one-to-one we can equate the exponents z + x = x and x + z = x. By
subtracting the real number x from both sides of each equation we find that z = 0
(since z + x − x = x + z − x = z and x − x = 0 for x, z ∈ R). Thus 0 = e0 for X with
vector addition and scalar multiplication as defined in the question. (Notice that the
zero vector does not depend on x, and therefore also not on a; otherwise it would not
satisfy axiom 4 in Definition 1 on page 172 of the textbook.)
1. Let a = ex ∈ X and b = ey ∈ X, i.e. x, y ∈ R. Then a ⊕ b = ex+y ∈ X since the

first value of the pair is 1, and the second x + y ∈ R is a real number.
2. Let a = ex ∈ X and b = ey ∈ X. Then
a ⊕ b = ex ⊕ ey = ex+y
b ⊕ a = ey ⊕ ex = ey+x
so that a ⊕ b = b ⊕ a (since the addition of real numbers is commutative, i.e.

x + y = y + x).
3. Let a = ex ∈ X, b = ey ∈ X and c = ez ∈ X. Then
(a ⊕ b) ⊕ c = ex+y ⊕ ez = e(x+y)+z = ex+y+z

a ⊕ (b ⊕ c) = ex ⊕ ey+z = ex+(y+z) = ex+y+z
due to the associativity of real numbers, i.e. (x + y) + z = x + (y + z). Thus

(a ⊕ b) ⊕ c = a ⊕ (b ⊕ c).
61
4. The existence of the zero vector is demonstrated in Question (4.1).
0
5. Let a = ex ∈ X. Assume there exists x0 ∈ R such that −a = ex , i.e.
0 0
a ⊕ (−a) = ex ⊕ ex = ex+x = e0 = 0
0 0
(−a) ⊕ a = ex ⊕ ex = ex +x = e0 = 0.
We can equate exponents, i.e. x + x0 = 0 and x0 + x = 0. Thus x0 = −x.

Consequently −a exists and is given by −a = −ex = e−x .
6. Let k ∈ R and a = ex ∈ X (so that x ∈ R). Then k a = ekx ∈ X since kx ∈ R

is a real number.
7. Let k ∈ R, a = ex ∈ X and b = ey ∈ X. Then
k (a ⊕ b) = k (ex ⊕ ey ) = k ex+y = ek(x+y) = ekx+ky

(k a) ⊕ (k b) = ekx ⊕ eky = ekx+ky
so that k (a ⊕ b) = k a⊕k b.
8. Let k, m ∈ R and a = ex ∈ X. Then
(k + m) a = e(k+m)x = ekx+mx
(k a) ⊕ (m a) = ekx ⊕ emx = ekx+mx
so that (k + m) a=k a⊕m a.
9. Let k, m ∈ R and a = ex ∈ X. Then
(km) a = e(km)x = ekmx

k (m a) = k emx = ek(mx) = ekmx
due to the associativity of products of real numbers. Thus (km) a = k (m a).
10. Let a = ex ∈ X. Then
1 a=1 ex = e1x = ex = a.
(4.3) Is the vector space X a subspace of R ? Motivate your answer.

A subspace X of a vector space V is a subset of V together with the same vector addition
and scalar multiplication as in V . Here we use a vector addition for X which is different
from the usual vector addition in R. Hence X is not a subspace of R.
62
MAT2611/101/3/2017
Question 5
Show that
Y := { (0, y) : y ∈ R },
with the usual vector addition and scalar multiplication in R2 , is a subspace of R2 .
• First, we show that Y is non-empty (although this is obvious). It is sufficient to check whether
the zero vector (0, 0) in R2 is also in Y (i.e. this condition together with the next two conditions
are necessary and sufficient for Y to be a vector subspace of R2 ) . Since the first number in the
pair is 0, and the second number 0 ∈ R we have (0, 0) ∈ Y . Thus Y is non-empty.
• Second, we show that Y is closed under the usual vector addition in R2 . Let a, b ∈ Y , i.e. there
exists x, y ∈ R such that a = (0, x) and b = (0, y). The usual vector addition in R2 yields
a + b = (0, x) + (0, y) = (0 + 0, x + y) = (0, x + y) ∈ Y
since x + y ∈ R.
• Third, we show that Y is closed under the usual scalar multiplication in R2 . Let k ∈ R and
a ∈ Y , i.e. there exists x ∈ R such that a = (0, x). The usual scalar multiplication in R2 yields
k · a = k · (0, x) = (k0, kx) = (0, kx) ∈ Y
since kx ∈ R.
Thus by Theorem 4.2.1 of the textbook, Y is a subspace of R2 .
Question 6
Let B ∈ M22 be a fixed but arbitrary 2 × 2 matrix. Show that

Y := { A : A ∈ M22 , AB = BA },
with the usual vector addition and scalar multiplication in M22 , is a subspace of M22 .
• First, we show that Y is non-empty (although this is obvious). It is sufficient to check whether
the zero vector [ 00 00 ] in M22 is also in Y (i.e. this condition together with the next two conditions
are necessary and sufficient for Y to be a vector subspace of M22 ). Since [ 00 00 ] B = B [ 00 00 ] = [ 00 00 ]
we have [ 00 00 ] ∈ Y . Thus Y is non-empty.
• Second, we show that Y is closed under the usual vector addition in M22 . Let G, H ∈ Y , i.e.
G, H ∈ M22 , GB = BG and HB = BH. The usual vector addition in M22 yields G + H ∈ M22
and
(G + H)B = GB + HB = BG + BH = B(G + H)
so that G + H ∈ Y .
• Third, we show that Y is closed under the usual scalar multiplication in R2 . Let k ∈ R and
G ∈ Y , i.e. G ∈ M22 and GB = BG. The usual scalar multiplication in M22 yields k · G ∈ M22
and
(k · G)B = k · (GB) = k · (BG) = B(k · G)
so that k · G ∈ Y .
Thus by Theorem 4.2.1 of the textbook, Y is a subspace of M22 .
63
Question 7
Is
1 2 1 1 2
span , , = span , ?
2 1 1 2 1
Noting that

1 1 2 2 1 2 1 1 1 1 2
=1 +0 =0 +1 = +
2 2 1 1 2 1 1 3 2 3 1

1 1 2 1 2 1 2 1
=1 +0 +0 =0 +1 +0
2 2 1 1 1 2 1 1
then by Theorem 4.2.5 of the textbook, the answer is yes.
Alternative:
To answer this question, we recall that for two sets A and B we have A = B if and only if A ⊆ B and
B ⊆ A. We have

1 2 1 1 2 1
span , , = a +b +c : a, b, c ∈ R
2 1 1 2 1 1
and
1 2 1 2
span , = α +β : α, β ∈ R .
2 1 2 1
We need to determine:
1. Given a, b, c ∈ R does α, β ∈ R exist such that

1 2 1 1 2
a +b +c =α +β ?
2 1 1 2 1
Thus we have to satisfy a + 2b + c = α + 2β and 2a + b + c = 2α + β. Solving for α and β yields
α = a + c/3 and β = b + c/3, i.e. we found a solution.
2. Given α, β ∈ R does a, b, c ∈ R exist such that

1 2 1 1 2
a +b +c =α +β ?
2 1 1 2 1
Thus we have to satisfy a + 2b + c = α + 2β and 2a + b + c = 2α + β. Solving for a, b and c
yields a = α − c/3, b = β − c/3 and c ∈ R is free, i.e. we found a solution (many solutions).
Thus the answer is yes.
Question 8
Is
1 2 1
span , ,
2 1 1
a linearly independent subset of R2 ? Motivate your answer.
Since
1 2 1 2 1 2 1
, , + ∈ span , ,
2 1 2 1 2 1 1
by Theorem 4.3.1(a) in the textbook, the given set is not a linearly independent subset of R2 . (By
this theorem, the span of any set of vectors is not linearly independent.)
64
MAT2611/101/3/2017
Question 9

3 6 0 −1 0 −8 1 0
A1 = ; A2 = ; A3 = ; and A4 = .
3 −6 −1 0 −12 −4 −1 2
First we show linear independence of the vectors. Let c1 , c2 , c3 , c4 ∈ R determined by
c1 A1 + c2 A2 + c3 A3 + c4 A4 = 0
i.e.

3c1 6c1 0 −c2 0 −8c3 c4 0
+ + +
3c1 −6c1 −c2 0 −12c3 −4c3 −c4 2c4

3c1 + c4 6c1 − c2 − 8c3
=
3c1 − c2 − 12c3 − c4 −6c1 − 4c3 + 2c4

0 0
=
0 0
Thus we obtain the four equations
3c1 + c4 =0
6c1 − c2 − 8c3 =0
3c1 − c2 − 12c3 − c4 =0
−6c1 − 4c3 + 2c4 =0
or in matrix form     
3 0 0 1 c1 0
 6 −1 −8 0  c2  0
 3 −1 −12 −1 c3  = 0 .
    
−6 0 −4 2 c4 0
Row reduction of the augmented matrix yields
   
3 0 0 1 0 3 0 0 1 0
 6 −1 −8 0 0  →
−2R1    0 −1 −8 −2 0 
 
−R1  3 −1 −12 −1 0  −R2  0 −1 −12 −2 0 
+2R1 −6 0 −4 2 0 0 0 −4 4 0
 
3 0 0 1 0
 0 −1 −8 −2 0 
→ 
 0 0 −4

0 0 
−R3 0 0 −4 4 0
 
3 0 0 1 0
 0 −1 −8 −2 0 
→ 
 0 0 −4

0 0 
0 0 0 4 0
65
Here −2R1 means subtract twice the first row from the second row (appearing right
of −2R1). Thus c4 = 0, c3 = 0, c2 = −8c3 − 2c4 = 0 and c1 = −c4 /3 = 0. Since
c1 = c2 = c3 = c4 = 0 is the only solution, the matrices A1 , A2 , A3 and A4 are linearly
independent.
Next we show that any element of M22 can be expressed as a linear combination of A1 ,
A2 , A3 and A4 . Let a, b, c, d ∈ R and

a b 3a1 + a4 6a1 − a2 − 8a3
= a1 A1 + a2 A2 + a3 A3 + a4 A4 =
c d 3a1 − a2 − 12a3 − a4 −6a1 − 4a3 + 2a4
3a1 + a4 =a
6a1 − a2 − 8a3 =b
3a1 − a2 − 12a3 − c4 =c
−6a1 − 4a3 + 2a4 =d
3 0 0 1 a1 a
 6 −1 −8 0  a2   b 
 3 −1 −12 −1 a3  =  c  .
    
−6 0 −4 2 a4 d
   
3 0 0 1 a 3 0 0 1 a
−2R1
 6 −1 −8 0 b  →  0 −1 −8 −2 b − 2a 
  

−R1  3 −1 −12 −1 c  −R2  0 −1 −12 −2 c − a 
+2R1 −6 0 −4 2 d 0 0 −4 4 d + 2a
 
3 0 0 1 a
 0 −1 −8 −2 b − 2a 
→  
 0 0 −4 0 a − b + c 
−R3 0 0 −4 4 d + 2a
 
3 0 0 1 a
 0 −1 −8 −2 b − 2a 
→  
 0 0 −4 0 a−b+c 
0 0 0 4 a+b+d−c
Thus we find
a+b+d−c
a4 =
4
b−a−c
a3 =
4
a+b+d−c 7a − 7b + 5c − d
a2 = −8a3 − 2a4 − b + 2a = −2b + 2a + 2c − − b + 2a =
2 2
a − a4 3a − b + c − d
a1 = =
3 12
Since a solution exists, we have span(B) = M22 . Thus B is a basis for M22 .
66
MAT2611/101/3/2017

6 2
5 3
From 9.1 we have (for a = 6, b = 2, c = 5 and d = 3)
3 9 3
a4 = , a3 = − , a2 = 25, a1 =
2 4 2
so that

6 2 3 3 6 0 −1 9 0 −8 3 1 0
= + 25 − + .
5 3 2 3 −6 −1 0 4 −12 −4 2 −1 2

D = {A ∈ M22 : AT + A = 0}
First we must show that D is not empty. Note that if D is a subspace of M22 then it
is a vector space with the same zero vector as for M22 (i.e. the 2 × 2 zero matrix). It
suffices to check whether the zero vector is in D:
T
0 0 0 0 0 0 0 0 0 0
+ = + =
0 0 0 0 0 0 0 0 0 0

0 0
so that ∈ D.
0 0
Next we prove the closure under vector addition and under scalar multiplication.
Let A ∈ D and k ∈ R. Then, using the properties of the transpose and vector space,
we have
(kA)T + (kA) = kAT + kA = k(AT + A) = k0 = 0
where 0 on the right hand side denotes the 2 × 2 zero matrix. Thus D is closed under
scalar multiplication.
Let A, B ∈ D. Then, using the properties of the transpose and vector space, we find
(A + B)T + (A + B) = (AT + B T ) + (A + B) = (AT + A) + (B T + B) = 0 + 0 = 0
vector addition.
67
Question 10

3 −2 1 −1 0 1 1 0
A1 = ; A2 = ; A3 = ; and A4 = .
0 1 0 0 1 0 0 0
First we show linear independence of the vectors. Let c1 , c2 , c3 , c4 ∈ R determined by
c1 A1 + c2 A2 + c3 A3 + c4 A4 = 0
i.e.

3c1 −2c1 c2 −c2 0 c3 c4 0
+ + +
0 c1 0 0 c3 0 0 0

3c1 + c2 + c4 −2c1 − c2 + c3
=
c3 c1

0 0
=
0 0
3c1 + c2 + c4 =0
−2c1 − c2 + c3 =0
c3 =0
c1 =0
3 1 0 1 c1 0
−2 −1 1 0 c2  0
   =  .
0 0 1 0 c3  0
1 0 0 0 c4 0
   
=R3 3 1 0 1 0 1 0 0 0 0
 −2 −1 1 0 0  +2R1  −2 −1 1 0 0 
 →  
 0 0 1 0 0   0 0 1 0 0 
=R1 1 0 0 0 0 −3R1 3 1 0 1 0
 
1 0 0 0 0
−R2  0 −1 1 0 0 
→ 
 0 0 1 0 0 

+R2 0 1 0 1 0
 
1 0 0 0 0
 0 1 −1 0 0 
→ 
 0 0 1 0 0 

−R3 0 0 1 1 0
 
1 0 0 0 0
 0 1 −1 0 0 
→ 
 0 0 1 0 0 

0 0 0 1 0
68
MAT2611/101/3/2017
Here = R3 means replace row 1 (appearing right of = R3 ) with row 3 and +2R1 means
add twice the first row from the second row (appearing right of +2R1 ). Thus c4 = 0,
c3 = 0, c2 = c3 = 0 and c1 = 0. Since c1 = c2 = c3 = c4 = 0 is the only solution, the
matrices A1 , A2 , A3 and A4 are linearly independent.
Next we show that any element of M22 can be expressed as a linear combination of A1 ,
A2 , A3 and A4 . Let a, b, c, d ∈ R and

a b 3a1 + a2 + a4 −2a1 − a2 + a3
= a1 A1 + a2 A2 + a3 A3 + a4 A4 =
c d a3 a1
3a1 + a2 + a4 =a
−2a1 − a2 + a3 =b
a3 =c
a1 =d
3 1 0 1 a1 a
−2 −1 1 0 a2   b 
   =  .
0 0 1 0 a3   c 
1 0 0 0 a4 d
   
=R3 3 1 0 1 a 1 0 0 0 d
 −2 −1 1 0 b   → +2R1  −2 −1 1 0 b 
 

 0 0 1 0 c   0 0 1 0 c 
=R1 1 0 0 0 d −3R1 3 1 0 1 a
 
1 0 0 0 d
−R2  0 −1 1 0 b + 2d 
→ 
 0 0 1 0

c 
+R2 0 1 0 1 a − 3d
 
1 0 0 0 d
 0 1 −1 0 −b − 2d 
→ 
 0 0 1 0

c 
−R3 0 0 1 1 a+b−d
 
1 0 0 0 d
 0 1 −1 0 −b − 2d 
→ 
 0 0 1 0

c 
0 0 0 1 a+b−c−d
Thus we find
a4 =a+b−c−d
a3 =c
a2 = −b − 2d + a3 = −b − 2d + c
a1 =d
Since a solution exists, we have span(B) = M22 . Thus B is a basis for M22 .
69

6 2
5 3
From 10.1 we have (for a = 6, b = 2, c = 5 and d = 3)
a4 = 0, a3 = 5, a2 = −3, a1 = 3
so that
6 2 3 −2 1 −1 0 1 1 0
=3 −3 +5 +0 .
5 3 0 1 0 0 1 0 0 0

D = {A ∈ M22 : tr(A) = 0}
First we must show that D is not empty. Note that if D is a subspace of M22 then it
is a vector space with the same zero vector as for M22 (i.e. the 2 × 2 zero matrix). It
suffices to check whether the zero vector is in D:

0 0
tr =0+0=0
0 0

0 0
so that ∈ D.
0 0
Let A ∈ D and k ∈ R. Then, using the properties of the trace and vector space, we
have
tr(kA) = k tr(A) = k0 = 0
scalar multiplication.
Let A, B ∈ D. Then, using the properties of the trace and vector space, we find
tr(A + B) = tr(A) + tr(B) = 0 + 0 = 0
vector addition.
70
MAT2611/101/3/2017
Question 11
Given the matrix

1 2 1
A=
2 1 1

The range of A is the column space of A

1 2 1
R(A) = a +b +c : a, b, c ∈ R (this is sufficient)
2 1 1

b + (a + b + c)
= : a, b, c ∈ R
a + (a + b + c)
0
b 0 0 (e.g. set c = −a − b,
= : a ,b ∈ R
a0 a = a0 , b = b 0 )
= R2 .

The rank of A is the dimension of the column space of A (and also the dimension of
the row space of A). Since the number of rows is 2, rank(A) ≤ 2. From Question
(12.1) it is clear that rank(A) = 2. Here we show an alternative method to obtain this
answer. The dimension of the row space is the number of non-zero rows obtained after
row reduction of A:

1 2 1 1 2 1
→ (R2 ← R2 − 2R1 )
2 1 1 0 −3 −1
Since the last matrix is in upper triangular form, no further row reduction steps are
required. There are two non-zero rows, consequently rank(A) = 2.
(11.3) Find the null space of A.

The null space of A is given by
    
 x x 
y  : x, y, z ∈ R, A y  = 0
0 
z z


1 2 1 : 0 1 2 1 : 0
→ (R2 ← R2 − 2R1 )
2 1 1 : 0 0 −3 −1 : 0

1 2 1 : 0
→ (R2 ← −R2 /3)
0 1 1/3 : 0

1 0 1/3 : 0
→ (R1 ← R1 − 2R2 )
0 1 1/3 : 0
71
so that the null space of A is
       
 x x   −1/3z
y  : x, y, z ∈ R, A y  = 0

= −1/3z : z ∈ R

0  
z z z
 
   
 −1/3 
= z −1/3 : z ∈ R
1
 

The nullity of A is the dimension of the null space of A (by Theorem 4.8.3: the number
of free parameters in Question (12.3)) which is 1. Alternatively, since rank(A) +
nullity(A) = 3 and rank(A) = 2 then nullity(A) = 1.
Question 12
Given the matrix  

1 2
A = 2 1
1 1

The range of A is the column space of A
     
 1 2 
R(A) = a 2 + b 1 : a, b ∈ R .
1 1
 

The rank of A is the dimension of the column space of A (and also the dimension of
the row space of A). Since the number of rows is 2, rank(A) ≤ 2. From Question
(12.1) it is clear that rank(A) = 2. Here we show an alternative method to obtain this
answer. The dimension of the row space is the number of non-zero rows obtained after
row reduction of A:
   
1 2 1 2
2 1 → 0 −1 (R2 ← R2 − 2R1 )
1 1 1 1
 
1 2
→ 0 −1 (R3 ← R3 − R1 )
0 −1
 
1 2
→ 0 −1 (R3 ← R3 − R2 )
0 0
Since the last matrix is in upper triangular form, no further row reduction steps are
required. There are two non-zero rows, consequently rank(A) = 2.
72
MAT2611/101/3/2017
(12.3) Find the nullspace of A.

The null space of A is given by
  
 0 
x x
: x, y ∈ R, A = 0

 y y
0


   
1 2 : 0 1 2 : 0
2 1 : 0 → 0 −1 : 0 (R2 ← R2 − 2R1 )
1 1 : 0 1 1 : 0
 
1 2 : 0
→ 0 −1 : 0 (R3 ← R3 − R1 )
0 −1 : 0
 
1 2 : 0
→ 0 −1 : 0 (R3 ← R3 − R2 )
0 0 : 0
 
1 2 : 0
→ 0 1 : 0 (R2 ← −R2 )
0 0 : 0
 
1 0 : 0
→ 0 1 : 0 (R1 ← R1 − 2R2 )
0 0 : 0
so that the null space of A is

  
 0 
x x 0
: x, y ∈ R, A = 0 = .
 y y 0
0


The nullity of A is the dimension of the null space of A (by Theorem 4.8.3: the number
of free parameters in Question (12.3)) which is 0. Alternatively, since rank(A) +
nullity(A) = 2 and rank(A) = 2 then nullity(A) = 0.
Question 13
(13.1) Let S and T be subspaces of a vector space V . Prove that S ∩ T is also a subspace of
V.
First we must show that S ∩ T is not empty. Note that if S, T and S ∩ T are subspaces
of V then they are vector spaces with the same zero vector 0 as for V . It suffices to
check whether the zero vector is in S ∩ T . Obviously 0 ∈ S and 0 ∈ T since S and T
are subspaces of V . Thus 0 ∈ S ∩ T .
73
Let A ∈ S ∩ T and k ∈ R. Then, using the properties of the vector spaces S and T , we
have A, kA ∈ S and A, kA ∈ T so that kA ∈ S ∩ T . Thus S ∩ T is closed under scalar
multiplication.
Let A, B ∈ S ∩ T . Consequently A, B ∈ S and A, B ∈ T . Using the properties of the

vector spaces S and T we find A + B ∈ S and A + B ∈ T so that A + B ∈ S ∩ T . Thus
S ∩ T is closed under vector addition.
It follows that S ∩ T is also a subspace of V .
(13.2) Let S = {(a + b, a, −a, −b) : a, b ∈ R} and T = {(x, 2y, −y, −x) : x, y ∈ R}.
Setting a = b = 0 we find that the zero vector of R4 is in S. Let a, b, k ∈ R and
(a + b, a, −a, −b) ∈ S, then
k(a + b, a, −a, −b) = (ka + kb, ka, −ka, −kb) = (a0 + b0 , a0 , −a0 , −b0 ) ∈ S
where a0 := ka and b0 := kb. Thus S is closed under scalar multiplication. Let

a1 , b1 , a2 , b2 ∈ R and (a1 + b1 , a1 , −a1 , −b1 ), (a2 + b2 , a2 , −a2 , −b2 ) ∈ S, then
(a1 + b1 , a1 , −a1 , −b1 ) + (a2 + b2 , a2 , −a2 , −b2 )

= (a1 + a2 + b1 + b2 , a1 + a2 , −a1 − a2 , −b1 − b2 )
= (a0 + b0 , a0 , −a0 , −b0 ) ∈ S
where a0 := a1 + a2 and b0 := b1 + b2 . Thus S is closed under vector addition. It

follows that S is a subspace of R4 .
Setting x = y = 0 we find that the zero vector of R4 is in T . Let k, x, y ∈ R and

(x, 2y, −y, −x) ∈ T , then
k(x, 2y, −y, −x) = (kx, 2ky, −ky, −kx) = (x0 , 2y 0 , −y 0 , −x0 ) ∈ T
where x0 := kx and y 0 := ky. Thus T is closed under scalar multiplication. Let

x1 , y1 , x2 , y2 ∈ R and (x1 , 2y1 , −y1 , −x1 ), (x2 , 2y2 , −y2 , −x2 ) ∈ T , then
(x1 , 2y1 , −y1 , −x1 ) + (x2 , 2y2 , −y2 , −x2 ) = (x1 + x2 , 2(y1 + y2 ), −y1 − y2 , −x1 − x2 )
= (x0 , 2y 0 , −y 0 , −x0 ) ∈ T
where x0 := x1 + x2 and y 0 := y1 + y2 . Thus T is closed under vector addition. It

follows that T is a subspace of R4 .

Note that elements of S can be written in the form
(a + b, a, −a, −b) = a(1, 1, −1, 0) + b(1, 0, 0, −1)
74
MAT2611/101/3/2017
and that
c1 (1, 1, −1, 0) + c2 (1, 0, 0, −1) = (0, 0, 0, 0)
has only the trivial solution c1 = c2 = 0 (from the third and fourth components).
The two vectors (1, 1, −1, 0) and (1, 0, 0, −1) are linearly independent, so the di-
mension of S is 2.
Note that elements of T can be written in the form
(x, 2y, −y, −x) = x(1, 0, 0, −1) + y(0, 2, −1, 0)
and that
c1 (1, 0, 0, −1) + c2 (0, 2, −1, 0) = (0, 0, 0, 0)
has only the trivial solution c1 = c2 = 0 (from the first and second components).
The two vectors (1, 0, 0, −1) and (0, 2, −1, 0) are linearly independent, so the di-
mension of T is 2.
Let a, b, x, y ∈ R and let (a + b, a, −a, −b) ∈ S and (x, 2y, −y, −x) ∈ T . The
intersection S ∩ T is given by a, b, x, y satisfying
(a + b, a, −a, −b) = (x, 2y, −y, −x).
The third component gives a = y while the second provides a = 2y = 2a so that

a = y = 0. The first component now gives a + b = b = x and the fourth component
−b = −x is identically satisfied. It follows that
S ∩ T = {(x, 0, 0, −x) : x ∈ R}.
We have one free parameter (x) describing S ∩ T . The dimension is 1. A basis is

  

 1 
  
0 .

  0 
 
−1
 
(13.3) Let
S = {1 − x; 5 + 3x − 2x2 ; 1 + 3x − x2 } ⊂ P2 .
There are different approaches to determine a basis. First we consider a straightforward
method given that B = {1, x, x2 } is a basis for P2 . Then we have the representations
(using the ordering 1, x, x2 )
     
1 5 1
2 2
[1 − x]B = −1
  [5 + 3x − 2x ]B = 3
  [1 + 3x − x ]B = 3 

0 −2 −1
75
Using these column vectors as rows in a matrix and applying row reduction
   
1 −1 0 1 −1 0
−5R1  5 3 −2  →  0 8 −2 
−R1 1 3 −1 −R2/2 0 4 −1
 
1 −1 0
→  0 8 −2 
0 0 0
we find the basis {1 − x; 8x − 2x2 } and dimension 2. Of course, other choices of basis
are possible.
Note: In the absence of an existing basis for P2 we need to first determine which subsets
are linearly independent and take (one of) the largest linearly independent subsets as
a basis. Since S is finite, we can do this by enumerating the subsets. First we try all
subsets with 3 elements (i.e. the set S):
0 = c11 (1 − x) + c12 (5 + 3x − 2x2 ) + c13 (1 + 3x − x2 ) = 0

0 = c11 + 5c12 + c13
0 = −c11 + 3c12 + 3c13
0 = −2c12 − c13
The last equation yields c13 = −2c12 . The second last equation becomes c11 = −3c12 .
For example, setting c12 = 1 so that c11 = −3 and c13 = −2 yields a nontrivial solution.
This set is not linearly independent.
Next we try all subsets with 2 elements until we find a linearly independent set.
First {1 − x; 5 + 3x − 2x2 }:
0 = c21 (1 − x) + c22 (5 + 3x − 2x2 ) = 0

0 = c21 + 5c22
0 = −c21 + 3c22
0 = c22
which has only the trivial solution c21 = c22 = 0. Thus a basis is {1 − x; 5 + 3x − 2x2 }
and the dimension is 2. (In fact, any two elements from S form a basis for span(S); test
this yourself.)
Question 14
(14.1) Let S = {(a, b, b, a + c) : a, b, c ∈ R} and T = {(x, y − x, y + z, z) : x, y, z ∈ R}.

Setting a = b = c = 0 we find that the zero vector of R4 is in S. Let a, b, c, k ∈ R
and (a, b, b, a + c) ∈ S, then
k(a, b, b, a + c) = (ka, kb, kb, ka + kc) = (a0 , b0 , b0 , a0 + c0 ) ∈ S
76
MAT2611/101/3/2017
where a0 := ka, b0 = kb and c0 := kc. Thus S is closed under scalar multiplication.

Let a1 , b1 , a2 , b2 , c1 , c2 ∈ R and (a1 , b1 , b1 , a1 + c1 ), (a2 , b2 , b2 , a2 + c2 ) ∈ S, then
(a1 , b1 , b1 , a1 + c1 ) + (a2 , b2 , b2 , a2 + c2 )
= (a1 + a2 , b1 + b2 , b1 + b2 , a1 + c1 + a2 + c2 )
= (a0 , b0 , b0 , a0 + c0 ) ∈ S
where a0 := a1 + a2 , b0 := b1 + b2 and c0 := c1 + c2 . Thus S is closed under vector
addition. It follows that S is a subspace of R4 .
Setting x = y = z = 0 we find that the zero vector of R4 is in T . Let k, x, y, z ∈ R

and (x, y − x, y + z, z) ∈ T , then
k(x, y − x, y + z, z) = (kx, ky − kx, ky + kz, kz) = (x0 , y 0 − x0 , y 0 + z 0 , z 0 ) ∈ T
where x0 := kx, y 0 = ky and z 0 := kz. Thus T is closed under scalar multiplication.
Let x1 , y1 , x2 , y2 , z1 , z2 ∈ R and (x1 , y1 − x1 , y1 + z1 , z1 ), (x2 , y2 − x2 , y2 + z2 , z2 ) ∈ T ,
then
(x1 , y1 − x1 , y1 + z1 , z1 ) + (x2 , y2 − x2 , y2 + z2 , z2 )
= (x1 + x2 , y1 − x1 + y2 − x2 , y1 + z1 + y2 + z2 , z1 + z2 )
= (x0 , y 0 − x0 , y 0 + z 0 , z 0 ) ∈ T
where x0 := x1 + x2 , y 0 ” = y1 + y2 and y 0 := y1 + y2 . Thus T is closed under vector
addition. It follows that T is a subspace of R4 .

Note that elements of S can be written in the form
(a, b, b, a + c) = a(1, 0, 0, 1) + b(0, 1, 1, 0) + c(0, 0, 0, 1)
and that
c1 (1, 0, 0, 1) + c2 (0, 1, 1, 0) + c3 (0, 0, 0, 1) = (0, 0, 0, 0)
has only the trivial solution c1 = c2 = c3 = 0 (from the third and fourth compo-
nents). The three vectors (1, 0, 0, 1), (0, 1, 1, 0) and (0, 0, 0, 1) are linearly indepen-
dent, so the dimension of S is 3.
Note that elements of T can be written in the form

(x, y − x, y + z, z) = x(1, −1, 0, 0) + y(0, 1, 1, 0) + z(0, 0, 1, 1)
and that
c1 (1, −1, 0, 0) + c2 (0, 1, 1, 0) + c3 (0, 0, 1, 1) = (0, 0, 0, 0)
has only the trivial solution c1 = c2 = c3 = 0 (from the first and second compo-
nents). The three vectors (1, −1, 0, 0), (0, 1, 1, 0) and (0, 0, 1, 1) are linearly inde-
pendent, so the dimension of T is 3.
77
Let a, b, c, x, y, z ∈ R and let (a, b, b, a + c) ∈ S and (x, y − x, y + z, z) ∈ T . The
intersection S ∩ T is given by a, b, x, y satisfying
(a, b, b, a + c) = (x, y − x, y + z, z)
⇒ a = x, b = y − x = y + z, a + c = z
⇒ a = x, b = y − a, z = −a, c = −2a.
It follows that
S ∩ T = {(a, y − a, y − a, −a) : a, y ∈ R}.
We have two free parameters (a and y) describing S ∩ T . The dimension is 2. A

basis is    

 1 0 
   
−1 , 1 .

 −1
   1

−1 0
 
(14.2) Let
S = {t3 + t2 − 2t + 1; t2 + 1; t3 − 2t; 2t3 + 3t2 − 4t + 3} ⊂ P3 .
There are different approaches to determine a basis. First we consider a straightforward
method given that B = {1, t, t2 , t3 } is a basis for P3 . Then we have the representations
(using the ordering 1, t, t2 , t3 )
       
1 1 0 3
 −2   0   −2   −4 
[t3 + t2 − 2t + 1]B = 
1
 [t2 + 1]B =   [t3 − 2t]B =   [2t3 + 3t2 − 4t + 3]B =  
1 0 3
1 0 1 2
Using these column vectors as rows in a matrix and applying row reduction
   
−R3 1 −2 1 1 1 0 1 0
 1 0 1 0  −R1
 1 0 1 0 
 

 0 −2 → 
0 1   0 −2 0 1 
3 −4 3 2 −3R1 3 −4 3 2
 
1 0 1 0
=R3
 0 0 0 0 

→ 
=R2  0 −2 0 1 
−2R3 0 −4 0 2
 
1 0 1 0
 0 −2 0 1 
→ 
 0

0 0 0 
0 0 0 0
we find the basis {1 + t2 , −2t + t3 } and dimension 2. Of course, other choices of basis
are possible.
78
MAT2611/101/3/2017
Question 15
p1 = 1 + 2x, p2 = 3 − x, q1 = 2 − 2x, q2 = 4 + 3x.
Solving
q1 = ap1 + bp2 ⇔ 2 − 2x = (a + 3b) + (2a − b)x ⇔ a + 3b = 2, 2a − b = −2
for a, b ∈ R yields a = −4/7 and b = 6/7. Solving
q2 = ap1 + bp2 ⇔ 4 + 3x = (a + 3b) + (2a − b)x ⇔ a + 3b = 4, 2a − b = 3
for a, b ∈ R yields a = 13/7 and b = 5/7. Thus

1 −4 13
PB 0 −→B = .
7 6 5
Solving
p1 = aq1 + bq2 ⇔ 1 + 2x = (2a + 4b) + (−2a + 3b)x ⇔ 2a + 4b = 1, −2a + 3b = 2
p2 = aq1 + bq2 ⇔ 3 − x = (2a + 4b) + (−2a + 3b)x ⇔ 2a + 4b = 3, −2a + 3b = −1

1 −5 13
QB−→B 0 = .
14 6 4
Note that

1 0
PB 0 −→B QB−→B 0 = QB−→B 0 PB 0 −→B = ⇔ QB−→B 0 = PB−10 −→B
0 1
so that an alternative method to find QB−→B 0 is to find PB−10 −→B .
(15.3) Compute [p]B 0 if p = 5 − x.

Solving
p = aq1 + bq2 ⇔ 5 − x = (2a + 4b) + (−2a + 3b)x ⇔ 2a + 4b = 5, −2a + 3b = −1

1 19
[p]B 0 = .
14 8
79
Question 16
p1 = 1 + 3x, p2 = 2 − x, q1 = 1 − 2x, q2 = 1 + x.
Solving
q1 = ap1 + bp2 ⇔ 1 − 2x = (a + 2b) + (3a − b)x ⇔ a + 2b = 1, 3a − b = −2
for a, b ∈ R yields b = 5/7 and a = −3/7. Solving
q2 = ap1 + bp2 ⇔ 1 + x = (a + 2b) + (3a − b)x ⇔ a + 2b = 1, 3a − b = 1
for a, b ∈ R yields b = 2/7 and a = 3/7. Thus

1 −3 3
PB 0 −→B = .
7 5 2
Solving
p1 = aq1 + bq2 ⇔ 1 + 3x = (a + b) + (−2a + b)x ⇔ a + b = 1, −2a + b = 3
p2 = aq1 + bq2 ⇔ 2 − x = (a + b) + (−2a + b)x ⇔ a + b = 2, −2a + b = −1
for a, b ∈ R yields a = 1 and b = 1. Thus

−2/3 1
QB−→B 0 = .
5/3 1
Note that

1 0
PB 0 −→B QB−→B 0 = QB−→B 0 PB 0 −→B = ⇔ QB−→B 0 = PB−10 −→B
0 1
so that an alternative method to find QB−→B 0 is to find PB−10 −→B .
(16.3) Compute [p]B 0 if p = 3 + x.

Solving
p = aq1 + bq2 ⇔ 3 + x = (a + b) + (−2a + b)x ⇔ a + b = 3, −2a + b = 1

1 2
[p]B 0 = .
3 7
80
MAT2611/101/3/2017
Question 17


First we show that span B2 = P1 :
span B2 = { a(1 + 2x) + b(2 + x) : a, b ∈ R } = { (a + 2b) + (2a + b)x : a, b ∈ R }
Now let α + βx ∈ P1 where α, β ∈ R are arbitrary. Solving α + βx = (a + 2b) + (2a + b)x

for a and b yields a + 2b = α and 2a + b = β so that a = (2β − α)/3 and b = (2α − β)/3,
i.e. α + βx ∈ span B2 . Since a + 2b, 2a + b ∈ R we also have (a + 2b) + (2a + b)x ∈ P1 .
Thus span B2 = P1 .
Next we show that B2 is a linearly independent set of vectors. Solving
c1 (1 + 2x) + c2 (2 + x) = (c1 + 2c2 ) + (2c1 + c2 )x = 0 + 0x
for c1 , c2 ∈ R yields c1 +2c2 = 0 and 2c1 +c2 = 0 (here we used the linear independence of
the standard basis {1, x} ≡ {1+0x, 0+1x} in P1 ). Thus c1 = −2c2 and c2 = −2c1 = 4c2
so that c1 = c2 = 0 is the only solution. Thus B2 is a linearly independent set of vectors.
Consequently B2 is a basis for P1 .
Alternatively, since B2 consists of two vectors (and P1 is 2-dimensional) by Theorem

4.5.4 of the textbook we need only show that B2 spans P1 or that B2 is linearly inde-
pendent, i.e. we can omit either half of the proof above.
Alternative:
The matrix representations of the elements of B2 with respect to the standard basis BP1
in P1 are
1 2
[1 + 2x]BP1 = , [2 + x]BP1 =
2 1
and taking the determinant of the matrix composed of these vectors as columns yields
1 2
= −3 6= 0
2 1
81
We express the elements of B1 in terms of B2 (using left to right ordering, i.e. 1 + x
first in B1 and 1 + 2x first in B2 ):
1 + x = p11 (1 + 2x) + p21 (2 + x) = (p11 + 2p21 ) + (2p11 + p21 )x

1 − x = p12 (1 + 2x) + p22 (2 + x) = (p12 + 2p22 ) + (2p12 + p22 )x
and solving for p11 , p12 , p21 , p22 ∈ R yields the two sets of equations (once again using
linear independence of the standard basis in P1 )
1 = p11 + 2p21 1 = p12 + 2p22

1 = 2p11 + p21 −1 = 2p12 + p22
These equations are straightforward to solve, p11 = 1/3, p21 = 1/3, p12 = −1 and
p22 = 1. The transition matrix is

p11 p12 1/3 −1 1 1 −3
PB1 →B2 = = = .
p21 p22 1/3 1 3 1 3
Alternative:
Using the fact that row reduction of the augmented matrix A : B leads to I : A−1 B

for square
matrices A and B with the same number of rows and A invertible, we have
that PB2 →B : PB1 →B reduces to
I : PB−1

P
2 →B B1 →B
= I : PB→B2 PB1 →B = I : PB1 →B2
for any basis B of P1 . Let B = BP1 = {1, x} be the standard basis in P1 , then

1 2
PB2 →B = [1 + 2x]BP1 [2 + x]BP1 = ,
2 1

1 1
PB1 →B = [1 + x]BP1 [1 − x]BP1 =
1 −1

1 2 : 1 1 1 2 : 1 1
→ (R2 ← R2 − 2R1 )
2 1 : 1 −1 0 −3 : −1 −3

1 2 : 1 1
→ (R2 ← −R2 /3)
0 1 : 1/3 1

1 0 : 1/3 −1
→ (R1 ← R1 − 2R2 ).
0 1 : 1/3 1
Thus
1/3 −1
PB1 →B2 = .
1/3 1
82
MAT2611/101/3/2017

1 1
PB2 →B3 = .
0 1

This is a change of basis from B1 to B3 , which can be achieved by changing basis
from B1 to B2 and then again changing basis from B2 to B3 (B1 → B3 ≡ B1 →
B2 → B3 ):

1 1 1 1 −3 1 2 0
PB1 →B3 = PB2 →B3 PB1 →B2 = = .
0 1 3 1 3 3 1 3

The columns of PB2 →B3 give B2 in terms of B3 . Let B3 = { p(x), q(x) } for some
p(x), q(x) ∈ P1 . The first column of PB2 →B3 provides

1
[1 + 2x]B3 = , 1 + 2x = 1p(x) + 0q(x) = p(x),
0
i.e. p(x) = 1 + 2x. The second column of PB2 →B3 provides

1
[2 + x]B3 = , 2 + x = 1p(x) + 1q(x) = p(x) + q(x) = 1 + 2x + q(x)
1
so that q(x) = 1 − x. Thus
B3 = { 1 + 2x, 1 − x }.
Exercise: verify this answer by calculating PB2 →B3 to see if the matrix you obtain
is the same as given in the question. Similarly, verify that PB1 →B3 is the same as
found in Question (18.3(a)).
Alternative: We use the columns of

1 −1
PB3 →B2 = PB−1
2 →B3
=
0 1
to find
B3 = { 1(1 + 2x) + 0(2 + x), −1(1 + 2x) + 1(2 + x) } = { 1 + 2x, 1 − x }.
83
Question 18


First we show that span B1 = P1 :
span B1 = { a(1 + x) + b(1 − x) : a, b ∈ R }P1 = { (a + b) + (a − b)x : a, b ∈ R }
Now let α + βx ∈ P1 where α, β ∈ R are arbitrary. Solving α + βx = (a + b) + (a − b)x

for a and b yields a + b = α and a − b = β so that a = (α + β)/2 and b = (α − β)/2,
i.e. α + βx ∈ span B1 . Since a + b, a − b ∈ R we also have (a + b) + (a − b)x ∈ P1 . Thus
span B1 = P1 .
Next we show that B1 is a linearly independent set of vectors. Solving
c1 (1 + x) + c2 (1 − x) = (c1 + c2 ) + (c1 − c2 )x = 0 + 0x
for c1 , c2 ∈ R yields c1 + c2 = 0 and c1 − c2 = 0 (here we used the linear independence of

the standard basis {1, x} ≡ {1 + 0x, 0 + 1x} in P1 ). Thus c1 = −c2 and c2 = c1 = −c2
so that c1 = c2 = 0 is the only solution. Thus B1 is a linearly independent set of vectors.
Alternatively, since B1 consists of two vectors (and P1 is 2-dimensional) by Theorem

4.5.4 of the textbook we need only show that B1 spans P1 or that B1 is linearly inde-
pendent, i.e. we can omit either half of the proof above.
Alternative:
The matrix representations of the elements of B1 with respect to the standard basis BP1
in P1 are
1 1
[1 + x]BP1 = , [1 − x]BP1 =
1 −1
and taking the determinant of the matrix composed of these vectors as columns yields
1 1
= −2 6= 0
1 −1
84
MAT2611/101/3/2017

We express the elements of B2 in terms of B1 (using left to right ordering, i.e. 1 + x
first in B1 and 1 + 2x first in B2 ):
1 + 2x = p11 (1 + x) + p21 (1 − x) = (p11 + p21 ) + (p11 − p21 )x

2 + x = p12 (1 + x) + p22 (1 − x) = (p12 + p22 ) + (p12 − p22 )x
and solving for p11 , p12 , p21 , p22 ∈ R yields the two sets of equations (once again using
linear independence of the standard basis in P1 )
1 = p11 + p21 2 = p12 + p22

2 = p11 − p21 −1 = p12 − p22
These equations are straightforward to solve, p11 = 3/2, p21 = −1/2, p12 = 3/2 and
p22 = 1/2. The transition matrix is

p11 p12 3/2 3/2 1 3 3
PB2 →B1 = = = .
p21 p22 −1/2 1/2 2 −1 1
Alternative:
Using the fact that row reduction of the augmented matrix A : B leads to I : A−1 B

for square
matrices A and B with the same number of rows and A invertible, we have
that PB1 →B : PB2 →B reduces to
I : PB−1

P
1 →B B2 →B
= I : PB→B1 PB2 →B = I : PB2 →B1
for any basis B of P1 . Let B = BP1 = {1, x} be the standard basis in P1 , then

1 1
PB1 →B = [1 + x]BP1 [1 − x]BP1 = ,
1 −1

1 2
PB2 →B = [1 + 2x]BP1 [2 + x]BP1 =
2 1

1 1 : 1 2 1 1 : 1 2
→ (R2 ← R2 − R1 )
1 −1 : 2 1 0 −2 : 1 −1

1 1 : 1 2
→ (R2 ← −R2 /2)
0 1 : −1/2 1/2

1 0 : 3/2 3/2
→ (R1 ← R1 − R2 ).
0 1 : −1/2 1/2
Thus
3/2 3/2
PB2 →B1 = .
−1/2 1/2
85
1 1
PB1 →B3 = .
0 2
This is a change of basis from B2 to B3 , which can be achieved by changing basis
from B2 to B1 and then again changing basis from B1 to B3 (B2 → B3 ≡ B2 →
B1 → B3 ):

1 1 1 3 3 1 2
PB2 →B3 = PB1 →B3 PB2 →B1 = = .
0 2 2 −1 1 −1 1
The columns of PB1 →B3 give B1 in terms of B3 . Let B3 = { p(x), q(x) } for some
p(x), q(x) ∈ P1 . The first column of PB1 →B3 provides

1
[1 + x]B3 = , 1 + x = 1p(x) + 0q(x) = p(x),
0
i.e. p(x) = 1 + x. The second column of PB1 →B3 provides

1
[1 − x]B3 = , 1 − x = 1p(x) + 2q(x) = p(x) + 2q(x) = 1 + x + 2q(x)
2
so that q(x) = −x. Thus
B3 = { 1 + x, −x }.
Exercise: verify this answer by calculating PB1 →B3 to see if the matrix you obtain
is the same as given in the question. Similarly, verify that PB2 →B3 is the same as
found in Question (18.3(a)).
Alternative: We use the columns of

1 −1/2
PB3 →B1 = PB−1
1 →B3
=
0 1/2
to find
B3 = { 1(1 + x) + 0(1 − x), −1/2(1 + x) + 1/2(1 − x) } = { 1 + x, −x }.
Question 19
1 2 3
A = 1 2 3 .
1 2 3

The characteristic equation is
    
1 0 0 1 2 3 λ − 1 −2 −3
det λ 0 1 0 − 1 2 3
      = −1 λ − 2 −3 = λ3 − 6λ2 = 0
0 0 1 1 2 3 −1 −2 λ − 3
86
MAT2611/101/3/2017

We solve the characteristic equation for λ:
λ3 − 6λ2 = 0 ⇒ λ2 (λ − 6) = 0
i.e. we find the eigenvalue λ = 0 (algebraic multiplicity 2) and the eigenvalue λ = 6

(algebraic multiplicity 1).
The eigenspace corresponding to the eigenvalue 0 is given by the solutions to the equation
           
1 0 0 1 2 3 x −1 −2 −3 x 0
0 0 1 0 − 1 2 3 y  = −1 −2 −3 y  = 0
0 0 1 1 2 3 z −1 −2 −3 z 0
where x, y, z ∈ R. Row reduction of the augmented matrix yields

   
−1 −2 −3 : 0 −1 −2 −3 : 0
−1 −2 −3 : 0 →  0 0 0 : 0 (R2 ← R2 − R1 , R3 ← R3 − R1 )
−1 −2 −3 : 0 0 0 0 : 0
so that x = −2y − 3z where y and z are free. The eigenspace corresponding to the
eigenvalue 0 is
        
 −2y − 3z   −2 −3 
 y  : y, z ∈ R = y 1 + z 0 : y, z ∈ R .
   
z 0 1
   
The dimension of this eigenspace is 2. The geometric multiplicity for the eigenvalue 0
is 2. A basis for the eigenspace is given by
   
 −2 −3 
 1 ,  0  .
0 1
 
Exercise: verify that the above vectors are linearly independent.
           
1 0 0 1 2 3 x 5 −2 −3 x 0
6 0 1 0 − 1 2 3 y  = −1 4 −3 y  = 0
0 0 1 1 2 3 z −1 −2 3 z 0
87
   
5 −2 −3 : 0 1 2 −3 : 0
−1 4 −3 : 0 → −1 4 −3 : 0 (R1 ← −R3 , R3 ← R1 )
−1 −2 3 : 0 5 −2 −3 : 0
 
1 2 −3 : 0
→ 0 6 −6 : 0 (R2 ← R2 + R1 , R3 ← R3 − 5R1 )
0 −12 12 : 0
 
1 2 −3 : 0
→ 0 1 −1 : 0 (R2 ← R2 /6, R3 ← R3 + 2R2 )
0 0 0 : 0
 
1 0 −1 : 0
→ 0 1 −1 : 0 (R1 ← R1 − 2R2 )
0 0 0 : 0
so that x = z and y = z where z is free. The eigenspace corresponding to the eigenvalue
6 is       
 z   1 
z  : z ∈ R = z 1 : z ∈ R .
z 1
   
 
 1 
1 .
1
 
Question 20
 
Consider the matrix 1 2 0
A = 2 1 0  .
0 0 −1

    
1 0 0 1 2 0 λ − 1 −2 0
det λ 0 1 0 − 2 1 0  = −2 λ − 1 0 = (λ + 1)((λ − 1)2 − 4) = 0.
0 0 1 0 0 −1 0 0 λ+1

We solve the characteristic equation for λ:
(λ + 1)((λ − 1)2 − 4) = (λ + 1)(λ + 1)(λ − 3) = 0
i.e. we find the eigenvalue λ = −1 (algebraic multiplicity 2) and the eigenvalue λ = 3
(algebraic multiplicity 1).
88
MAT2611/101/3/2017
The eigenspace corresponding to the eigenvalue −1 is given by the solutions to the
equation
           
1 0 0 1 2 0 x −2 −2 0 x 0
−1 0 1 0 − 2 1 0  y  = −2 −2 0 y  = 0
0 0 1 0 0 −1 z 0 0 0 z 0

   
−2 −2 0 : 0 −2 −2 0 : 0
−2 −2 0 : 0 → 0
  0 0 : 0 (R2 ← R2 − R1 )
0 0 0 : 0 0 0 0 : 0
so that x = −y where y and z are free. The eigenspace corresponding to the eigenvalue
−1 is         
 −y   −1 0 
 y  : y, z ∈ R = y  1  + z 0 : y, z ∈ R .
z 0 1
   
The dimension of this eigenspace is 2. The geometric multiplicity for the eigenvalue −1
   
 −1 0 
 1  , 0 .
0 1
 
Exercise: verify that the above vectors are linearly independent.
           
1 0 0 1 2 0 x 2 −2 0 x 0
3 0 1 0 − 2 1 0  y  = −2 2 0 y  = 0
0 0 1 0 0 −1 z 0 0 4 z 0

   
2 −2 0 : 0 2 −2 0 : 0
−2 2 0 : 0 → 0 0 0 : 0 (R1 ← R2 + R1 )
0 0 4 : 0 0 0 4 : 0
 
2 −2 0 : 0
→ 0 0 4 : 0 (R2 ↔ R3 )
0 0 0 : 0
so that x = y and z = 0 where y is free. The eigenspace corresponding to the eigenvalue

3 is       
 y   1 
y  : y ∈ R = y 1 : y ∈ R .
0 0
   
89
 
 1 
1 .
0
 
Question 21
Let  
−1 0 1
A= 0 2 0
0 −3 1

First we need the eigenvalues an eigenvectors of A. The characteristic equation is
  
λ 0 0 λ+1 0 −1
det   0 λ 0 −A = 0
  λ−2 0 = (λ + 1)(λ − 1)(λ − 2) = 0.
0 0 λ 0 3 λ−1
Thus the eigenvalues are −1 , 1 and 2 . Now we need to determine whether the
corresponding eigenvectors are linearly independent. However, all the eigenvalues are
different (i.e. each has multiplicity 1) so that linear independence follows directly and
we conclude that A is diagonalizable.

We determine the eigenvectors for the eigenvalue −1:
             
1 0 0 −1 0 1 x 0 0 −1 x −z 0
− 0 1 0 −  0 2 0   y = 0 −3 0
    y =
  −3y  = 0

0 0 1 0 −3 1 z 0 3 −2 z 3y − 2z 0
Obviously z = y = 0. Thus the eigenvectors corresponding to the eigenvalue -1 are

  
 x 
 0  : x ∈ R \ {0} .
0
 
T
A representative eigenvector is (any one could have been chosen) 1 0 0 . Check
this by multiplying the vector on the left by A.
We determine the eigenvectors for the eigenvalue 1:

            
1 0 0 −1 0 1 x 2 0 −1 x 2x − z 0
0 1 0 −  0 2 0   y = 0 −1 0
    y =
  −y  = 0

0 0 1 0 −3 1 z 0 3 0 z 3y 0
90
MAT2611/101/3/2017
Obviously y = 0 and z = 2x. Thus the eigenvectors corresponding to the eigenvalue 1

are   
 x 
 0  : x ∈ R \ {0} .
2x
 
T
A representative eigenvector is (any one could have been chosen) 1 0 2 . Check this
by multiplying the vector on the left by A.

             
1 0 0 −1 0 1 x 3 0 −1 x 3x − z 0
2 0 1 0 − 0
  2 0   y = 0
  0 0   y =
  0  = 0

0 0 1 0 −3 1 z 0 3 1 z 3y + z 0
Obviously z = 3x = −3y. Thus the eigenvectors corresponding to the eigenvalue 2 are

  
 x 
−x : x ∈ R \ {0} .
3x
 
T
A representative eigenvector is (any one could have been chosen) 1 −1 3 . Check
Note that each eigenspace has dimension 1. If the eigenspace had dimension greater than
1, we would choose that many linearly independent vectors from that eigenspace.
Using the representative vectors above we find

 
1 1 1
P = 0
 0 −1 .
0 2 3
The matrix D is given by  

−1 0 0
D =  0 1 0 .
0 0 2
The order of the eigenvalues on the diagonal corresponds to the order of the eigenvectors
as columns in P .
P −1 can be found by row reduction of the matrix P augmented with the 3 × 3 identity
91
matrix
   
1 1 1 1 0 0 1 1 1 1 0 0
=R3/2  0 0 −1 0 1 0  → −3R3/2  0 1 3/2 0 0 1/2 
=−R2 0 2 3 0 0 1 0 0 1 0 −1 0
 
−R2−R3 1 1 1 1 0 0
→  0 1 0 0 3/2 1/2 
0 0 1 0 −1 0
 
1 0 0 1 −1/2 −1/2
→  0 1 0 0 3/2 1/2 
0 0 1 0 −1 0
Thus  
1 −1/2 −1/2
P −1 = 0 3/2 1/2  .
0 −1 0
Remember to check for calculation errors by ensuring that P −1 AP = D.
Since P −1 AP = D we have P DP −1 = A so that A2 = P DP −1 P DP −1 = P D2 P −1 and
A3 = AA2 = P DP −1 P D2 P −1 = P D3 P −1 etc. Thus
  11  
1 1 1 −1 0 0 1 −1/2 −1/2
A11 = P D11 P −1 = 0 0 −1  0 1 0 0 3/2 1/2 
0 2 3 0 0 2 0 −1 0
   
1 1 1 −1 0 0 1 −1/2 −1/2
= 0 0 −1  0 1 0  0 3/2 1/2 
0 2 3 0 0 2048 0 −1 0
 
−1 −2046 1
= 0 2048 0
0 −6141 1
Question 22
Let  
1 5 0
A = 5 1 0
1 −1 6

First we need the eigenvalues an eigenvectors of A. The characteristic equation is
  
λ 0 0 λ − 1 −5 0
det  0 λ 0  − A = −5 λ − 1 0
0 0 λ −1 1 λ−6
= (λ − 1)2 (λ − 6) − 25(λ − 6)
= ((λ − 1)2 − 25)(λ − 6) = (λ − 6)2 (λ + 4) = 0.
92
MAT2611/101/3/2017
Thus the eigenvalues are 6 (twice) and −4 . Now we need to determine whether
three corresponding linearly independent eigenvectors can be found. We determine the
eigenvectors for the eigenvalue −4:
             
1 0 0 1 5 0 x −5 −5 0 x −5x − 5y 0
−4 0 1 0 −  5 1 0 y  = −5 −5 0  y  =  −5x − 5y  = 0
0 0 1 −1 1 6 z 1 −1 −10 z −x + y − 10z 0
Obviously x = −y = −5z. Thus the eigenvectors corresponding to the eigenvalue -4 are

  
 −5z 
 5z  : z ∈ R \ {0} .
z
 
T
A representative eigenvector is (any one could have been chosen) −5 5 1 . Check

             
1 0 0 1 5 0 x 5 −5 0 x 5x − 5y 0
6 0 1 0 − 5 1 0 y  = −5 5 0 y  = −5x + 5y  = 0
0 0 1 1 −1 6 z 1 −1 0 z x−y 0
Obviously x = y. (No restriction on z.) Thus the eigenvectors corresponding to the

eigenvalue 6 are   
 x 
x : (x, z) ∈ R2 \ {(0, 0)} .
z
 
Two linearly independent representative eigenvectors are (any two could have been cho-
T T
sen) 1 1 0 and 0 0 1 . Check this by multiplying the vector on the left by A.
Thus we found three linearly independent eigenvectors and A is diagonalizable.
(It is not necessary to check that eigenvectors corresponding to different eigenvalues are
linearly independent, they must necessarily be.)

 
−5 1 0
P =  5 1 0 .
1 0 1
The matrix D is given by  

−4 0 0
D =  0 6 0 .
0 0 6
93
The order of the eigenvalues on the diagonal corresponds to the order of the eigenvectors
as columns in P .
matrix
   
+5R3 −5 1 0 1 0 0 =R3 0 1 5 1 0 5
+R1  5 1 0 0 1 0 → =R2 /2  0 2 0 1 1 0 
1 0 1 0 0 1 =R1 1 0 1 0 0 1
 
1 0 1 0 0 1
→  0 1 0 1/2 1/2 0 
−R2 0 1 5 1 0 5
 
−R3 /5 1 0 1 0
0 1
→  0 1 0 1/2
1/2 0 
=R3 /5 −1/2 5
0 0 5 1/2
 
1 0 1 −1/10 1/10 0
→  0 1 0 1/2 1/2 0 
0 0 5 1/10 −1/10 1
Thus  
−1/10 1/10 0
P −1 =  1/2 1/2 0 .
1/10 −1/10 1
Remember to check for calculation errors by ensuring that P −1 AP = D.

Since P −1 AP = D we have P DP −1 = A so that A2 = P DP −1 P DP −1 = P D2 P −1 and
A3 = AA2 = P DP −1 P D2 P −1 = P D3 P −1 etc. Thus
  10  
−5 1 0 −4 0 0 −1/10 1/10 0
A10 = P D10 P −1 =  5 1 0  0 6 0  1/2 1/2 0
1 0 1 0 0 6 1/10 −1/10 1
   
−5 1 0 1048576 0 0 −1/10 1/10 0
=  5 1 0  0 60466176 0   1/2 1/2 0
1 0 1 0 0 60466176 1/10 −1/10 1
 
30757376 29708800 0
= 29708800 30757376
 0 .
5941760 −5941760 60466176
Question 23

1 2 3
A = 1 2 3 .
1 2 3
94
MAT2611/101/3/2017
(23.1) Find an invertible matrix P such that D := P −1 AP is diagonal. Determine D.

The eigenvalues of A are 0 (twice) and 6. A basis for the eigenspace corresponding to
the eigenvalue 0 is given by    
 −2 −3 
 1 ,  0  .
0 1
 
A basis for the eigenspace corresponding to the eigenvalue 6 is given by

 
 1 
1 .
1
 
Thus we may choose

   
−2 −3 1 0 0 0
P = 1 0 1 , D = 0 0 0 .
0 1 1 0 0 6

Since D is diagonal, the rank is the number of non-zero diagonal entries of D (i.e. the
number of non-zero eigenvalues of A). Thus rank(D) = rank(A) = 1.
Since  
0 0 0
Dn = 0 0 0 
0 0 6n
we have
      
0 0 0 −2 −3 1 0 0 0 −1 4 −3
1
An = P 0 0 0  P −1 =  1 0 1 0 0 0  −1 −2 3 
6
0 0 6n 0 1 1 0 0 6n 1 2 3
  
−2 −3 1 0 0 0
= 1 0 1  0 0 0 
n−1 n−1
0 1 1 6 2·6 3 · 6n−1
  
−2 −3 1 0 0 0
n−1 
=6 1 0 1   0 0 0
0 1 1 1 2 3
 
1 2 3
n−1 
=6 1 2 3 = 6n−1 A
1 2 3
where we used the Adjoint to find P −1 . Alternatively, we arrive at the same solution
   
0 0 0 0 0 0
n
A =P 0 0 0  P −1 = 6n−1 P 0 0 0 P −1 = 6n−1 A
0 0 6n 0 0 6
without determining P −1 .
95
(23.4) Show that  
1
A = 1 1 2 3 .

1
Use this expression for A to calculate A2 , A3 etc. and compare with your answer to
(24.3).
Straightforward matrix multiplication yields
   
1 1 2 3
1 1 2 3 = 1 2 3 = A.
1 1 2 3
We have
     
1 1 1
2
A2 = 62−1 A

A = 1 1 2 3 1 1 2 3 = 1 6 1 2 3 = 6A
    
1 1 1
A3 = (A2 )A = 6A2 = 6(6A) = 62 A A3 = 63−1 A
.. ..
. .
A = (An−1 )A = 6n−2 A2 = 6n−2 (6A) = 6n−1 A
n
A = 6n−1 A.
n
Question 24

1 2 0
A = 2 1 0  .
0 0 −1
(24.1) Find an orthogonal matrix P such that D := P T AP is diagonal. Determine D.

The eigenvalues of A are -1 (twice) and 3. A basis for the eigenspace corresponding to
the eigenvalue -1 is given by    
 −1 0 
 1  , 0 .
0 1
 

 
 1 
1 .
0
 
Notice that all three basis elements above are already pairwise orthogonal, thus we need
only divide each basis element by its Euclidean norm. Thus we may choose
 1
− √2 0 √12
  
−1 0 0
P =  √12 0 √12  , D =  0 −1 0 .
0 1 0 0 0 3
96
MAT2611/101/3/2017

Since D is diagonal, the rank is the number of non-zero diagonal entries of D (i.e. the
number of non-zero eigenvalues of A). Thus rank(D) = rank(A) = 3.

Since  
(−1)n 0 0
Dn =  0 (−1)n 0 
0 0 3n
we have
 1
√1
 1
− √2 0 − √2 √1 0
   
(−1)n 0 0 2 (−1)n 0 0 2
n T
A =P  0 (−1)n
0 P = √12
  0 √1  
2
0 (−1) n
0   0 0 1
1 √1
0 0 3n 0 1 0 0 0 3 n √ 0
 1   (−1)n+1 (−1)n  2 2
1
− √2 0 √2 √ √ 0
2 2
=  √1 1
0 √2   0 0 (−1)n 
 
2 n n
3 3
0 1 0 √
2
√
2
0
 n
3 + (−1)n 3n − (−1)n

0 
 3n − 2(−1)n 3n + 2(−1)n

=

0

2 2
 
n
0 0 (−1)
(24.4) Let B be a m × m matrix where m ∈ N, I be the m × m identity matrix and k ∈ R.

(a) Let x be an eigenvector of B with corresponding eigenvalue λ. Show that x is an
eigenvector of B + kI. What is the corresponding eigenvalue of B + kI?
We have
(B + kI)x = Bx + kIx = λx + kx = (λ + k)x.
Since x 6= 0 it follows that x is an eigenvector of B + kI with corresponding
eigenvalue λ + k.
(b) Assume that B is diagonalizable, is B + kI diagonalizable?
Since B is diagonalizable, there exists an invertible m × m matrix Q such that
Q−1 BQ is diagonal. Thus
Q−1 (B + kI)Q = Q−1 BQ + kQ−1 IQ = Q−1 BQ + kI
is diagonal (since the sum of diagonal matrices is diagonal). Consequently, B + kI

is diagonalizable.
97
Question 25
(25.1) Show that

   
x1 y1
hx, yi := 3x1 y1 + x2 y2 + x3 y3 , x = x2  , y = y2  ∈ R3
x3 y3
We have for k ∈ R and
     
x1 y1 z1
x = x2  , y = y2  , z = z2  ∈ R3
x3 y3 z3
1. hx, yi = 3x1 y1 + x2 y2 + x3 y3 = 3y1 x1 + y2 x2 + y3 x3 = hy, xi
2. hx + z, yi = 3(x1 + z1 )y1 + (x2 + z2 )y2 + (x3 + z3 )y3

= 3x1 y1 + 3z1 y1 + x2 y2 + z2 y2 + x3 y3 + z3 y3
= 3x1 y1 + x2 y2 + x3 y3 + 3z1 y1 + z2 y2 + z3 y3 = hx, yi + hz, yi
3. hkx, yi = 3(kx1 )y1 + (kx2 )y2 + (kx3 )y3 = k(3x1 y1 + x2 y2 + x3 y3 ) = khx, yi
4. hx, xi = 3x21 + x22 + x23 ≥ 0 so that hx, xi ≥ 0 and hx, xi = 0 if and only if
x1 = x2 = x3 = 0 (since x21 , x22 , x23 ≥ 0), i.e. x = 0.

     
 1 −1 −1 
0 ,  0  ,  1 
1 1 1
 
Let      
1 −1 −1
u1 := 0 , u2 :=  0  , u3 :=  1  .
1 1 1
98
MAT2611/101/3/2017
Then the Gram-Schmidt process provides

 
1
v1 := u1 = 0

1
hv1 , v1 i = 3(12 ) + 02 + 12 = 4
hu2 , v1 i = 3(−1)1 + 0 + 1 = −2
     
−1 1 −1
hu2 , v1 i 2 1
v2 := u2 − v1 =  0  + 0 =  0 
hv1 , v1 i 4 2
1 1 3
1
hv2 , v2 i = (3(−1)2 + 02 + 32 ) = 3
4
hu3 , v1 i = (3(−1)1 + 0 + 12 ) = −2
1
hu3 , v2 i = (3(−1)(−1) + 0 + 3) = 3
2
hu3 , v1 i hu3 , v2 i
v3 := u3 − v1 − v2
hv1 , v1 i hv2 , v2 i
       
−1 1 −1 0
2  3 1   
= 1 +
  0 − · 0 = 1 .
4 3 2
1 1 3 0
Thus we have the orthogonal basis (with respect to the inner product defined in 2.1)
     
 1 −1 0 
0 , 1  0  , 1 .
2
1 3 0
 
Normalizing these vectors yields the orthonormal basis

( )  1  
−1
 
0 
1 1 1  1  1
p v1 , p v2 , p v3 = 0 , √  0 , 1 .
hv1 , v1 i hv2 , v2 i hv3 , v3 i 2 2 3 3
1 0

Question 26
(26.1) Show that

   
x1 y1
  
x3 y3
     
x1 y1 z1
x = x2 , y = y2 , z = z2  ∈ R3
    
x3 y3 z3
99
1. hx, yi = x1 y1 + 3x2 y2 + x3 y3 = y1 x1 + 3y2 x2 + y3 x3 = hy, xi
2. hx + z, yi = (x1 + z1 )y1 + 3(x2 + z2 )y2 + (x3 + z3 )y3

= x1 y1 + z1 y1 + 3x2 y2 + 3z2 y2 + x3 y3 + z3 y3
= x1 y1 + 3x2 y2 + x3 y3 + z1 y1 + 3z2 y2 + z3 y3 = hx, yi + hz, yi
3. hkx, yi = (kx1 )y1 + 3(kx2 )y2 + (kx3 )y3 = k(x1 y1 + 3x2 y2 + x3 y3 ) = khx, yi
4. hx, xi = x21 + 3x22 + x23 ≥ 0 so that hx, xi = 0 hx, xi ≥ 0 and if and only if
x1 = x2 = x3 = 0 (since x21 , x22 , x23 ≥ 0), i.e. x = 0.

     
 1 −1 −1 
0 ,  0  ,  1 
1 1 1
 
Let      
1 −1 −1
u1 := 0 , u2 :=  0  , u3 :=  1  .
1 1 1
 
1
v1 := u1 = 0
1
hv1 , v1 i = 12 + 3 · 02 + 12 = 2
     
−1 1 −1
hu2 , v1 i 0   
v2 := u2 − v1 = 0 −
  0 = 0
hv1 , v1 i 2
1 1 1
hv2 , v2 i = (−1)2 + 3 · 02 + 12 = 2
hu3 , v1 i hu3 , v2 i
v3 := u3 − v1 − v2
hv1 , v1 i hv2 , v2 i
     
−1 1 −1
0 2
=  1  − 0 −  0 
2 2
1 1 1
 
0
= 1 .
0
Thus we have the orthogonal basis

     
 1 −1 0 
0 ,  0  , 1 .
1 1 0
 
100
MAT2611/101/3/2017
Normalizing these vectors yields the orthonormal basis

( )   
1
 
−1
 
0 
1 1 1  1 1 1
p v1 , p v2 , p v3 = √ 0 , √  0  , √ 1 .
hv1 , v1 i hv2 , v2 i hv3 , v3 i  2
1 2 1 3 0 
Question 27
(27.1) Show that

hu, vi = u1 v1 + 2u2 v2 + 3u3 v3
x = (x1 , x2 , x3 ), y = (y1 , y2 , y3 ), z = (z1 , z2 , z3 ) ∈ R3
1. hx, yi = x1 y1 + 2x2 y2 + 3x3 y3 = y1 x1 + 2y2 x2 + 3y3 x3 = hy, xi
2. hx + z, yi = (x1 + z1 )y1 + 2(x2 + z2 )y2 + 3(x3 + z3 )y3

= x1 y1 + z1 y1 + 2x2 y2 + 2z2 y2 + 3x3 y3 + 3z3 y3
= x1 y1 + 2x2 y2 + 3x3 y3 + z1 y1 + 2z2 y2 + 3z3 y3 = hx, yi + hz, yi
3. hkx, yi = (kx1 )y1 + 2(kx2 )y2 + 3(kx3 )y3 = k(x1 y1 + 2x2 y2 + 3x3 y3 ) = khx, yi
4. hx, xi = x21 + 2x22 + 3x23 ≥ 0

so that hx, xi = 0 if and only if x1 = x2 = x3 = 0 (since x21 , x22 , x23 ≥ 0) , i.e. x = 0.
(27.2) Let u = (1, 1, 1), v = (1, 1, 0) and w = (1, 0, 0). Show that B = {u, v, w} is linearly
Solving
c1 (1, 1, 1) + c2 (1, 1, 0) + c3 (1, 0, 0) = (c1 + c2 + c3 , c1 + c2 , c1 ) = (0, 0, 0)
for c1 , c2 , c3 we find c1 = 0 from the third component, and consequently c2 = 0 from

the second component and finally c3 = 0 This is the only solution, so that u, v and w
are linearly independent. Let x, y, z ∈ R. Solving
c1 (1, 1, 1) + c2 (1, 1, 0) + c3 (1, 0, 0) = (c1 + c2 + c3 , c1 + c2 , c1 ) = (x, y, z)
for c1 , c2 , c3 we find c1 = z from the third component, and consequently c2 = y − z

from the second component and finally c3 = x − y. Since we found a solution we have
span{u, v, w} = R3 .
101
Let
u1 := u = (1, 1, 1), u2 := v = (1, 1, 0), u3 := w = (1, 0, 0).
v10 := u1 = (1, 1, 1)
v0 1
v1 := p 01 0 = √ (1, 1, 1) where we used hv10 , v10 i = 6
hv1 , v1 i 6
1 1
v20 := u2 − hu2 , v1 iv1 = (1, 1, 0) − (1, 1, 1) = (1, 1, −1)
2 2
v20 1 3
v2 := p 0 0 = √ (1, 1, −1) where we used hv20 , v20 i =
hv2 , v2 i 6 2
v30 := u3 − hu3 , v1 iv1 − hu3 , v2 iv2
1 1 1
= (1, 0, 0) − (1, 1, 1) − (1, 1, −1) = (2, −1, 0)
6 6 3
0
v 1 6
v3 := p 03 0 = √ (2, −1, 0) where we used hv30 , v30 i =
hv3 , v3 i 6 9
Thus we have the orthonormal basis

1 1 1
√ (1, 1, 1), √ (1, 1, −1), √ (2, −1, 0) .
6 6 6

4 3
u= , 0, − and v = (0, 1, 0).
5 5

Notice that hu, vi = 0 and hu, ui = hv, vi = 1. Thus {u, v} is an orthonormal basis for
W . (If this was not the case, we could apply Gram-Schmidt to find such an orthonormal
basis). It follows that

4 3
w1 = hw, uiu + hw, viv = −u + 2v = − , 2,
5 5

9 12
w2 = w − w1 = , 0,
5 5

4 3 9 12
w = − , 2, + , 0, .
5 5 5 5
102
MAT2611/101/3/2017
Question 28
(28.1) Show that

hu, vi = 2u1 v1 + 3u2 v2 + u3 v3
x = (x1 , x2 , x3 ), y = (y1 , y2 , y3 ), z = (z1 , z2 , z3 ) ∈ R3
1. hx, yi = 2x1 y1 + 3x2 y2 + x3 y3 = 2y1 x1 + 3y2 x2 + y3 x3 = hy, xi
2. hx + z, yi = 2(x1 + z1 )y1 + 3(x2 + z2 )y2 + (x3 + z3 )y3

= 2x1 y1 + 2z1 y1 + 3x2 y2 + 3z2 y2 + x3 y3 + z3 y3
= 2x1 y1 + 3x2 y2 + x3 y3 + 2z1 y1 + 3z2 y2 + z3 y3 = hx, yi + hz, yi
3. hkx, yi = 2(kx1 )y1 + 3(kx2 )y2 + (kx3 )y3 = k(2x1 y1 + 3x2 y2 + x3 y3 ) = khx, yi
4. hx, xi = 2x21 + 3x22 + x23 ≥ 0

so that hx, xi = 0 if and only if x1 = x2 = x3 = 0 (since x21 , x22 , x23 ≥ 0) , i.e. x = 0.
(28.2) Let u = (1, 1, 1), v = (−1, 1, 0) and w = (1, 2, 1). Show that B = {u, v, w} is linearly
Solving
c1 (1, 1, 1) + c2 (−1, 1, 0) + c3 (1, 2, 1) = (c1 − c2 + c3 , c1 + c2 + 2c3 , c1 + c3 ) = (0, 0, 0)
for c1 , c2 , c3 and noting that for the determinant of the coefficient matrix
 
1 −1 1
det 1 1 2 = −1
1 0 1
we have a unique solution c1 = c2 = c3 = 0. Thus u, v and w are linearly independent.

Let x, y, z ∈ R. Solving
c1 (1, 1, 1) + c2 (−1, 1, 0) + c3 (1, 2, 1) = (c1 − c2 + c3 , c1 + c2 + 2c3 , c1 + c3 ) = (x, y, z)
for c1 , c2 , c3 and noting that for the determinant of the coefficient matrix
 
1 −1 1
det 1 1 2 = −1
1 0 1
we have a unique solution. Since we found a solution we have span{u, v, w} = R3 .
103
Let
u1 := u = (1, 1, 1), u2 := v = (−1, 1, 0), u3 := w = (1, 2, 1).
v10 := u1 = (1, 1, 1)
hv10 , v10 i = 2 · 1 · 1 + 3 · 1 · 1 + 1 · 1 = 6
v0 1
v1 := p 01 0 = √ (1, 1, 1)
hv1 , v1 i 6
1 1
hu2 , v1 i = √ (2 · (−1) · 1 + 3 · 1 · 1 + 0 · 1) = √
6 6
1 1
v20 := u2 − hu2 , v1 iv1 = (−1, 1, 0) − (1, 1, 1) = (−7, 5, −1)
6 6
0 0 1 29
hv2 , v2 i = (2 · (−7) · (−7) + 3 · 5 · 5 + (−1) · (−1)) =
36 6
v20 1
v2 := p 0 0 = √ √ (−7, 5, −1)
hv2 , v2 i 6 29
1 9
hu3 , v1 i = √ (2 · 1 · 1 + 3 · 2 · 1 + 1 · 1) = √
6 6
1 15
hu3 , v2 i = √ √ (2 · 1 · (−7) + 3 · 2 · 5 + 1 · (−1)) = √ √
6 29 6 29
0
v3 := u3 − hu3 , v1 iv1 − hu3 , v2 iv2
9 15 1
= (1, 2, 1) − (1, 1, 1) − (−7, 5, −1) = (3, 2, −12)
6 174 29
1 6
hv30 , v30 i = 2
(2 · 3 · 3 + 3 · 2 · 2 + (−12) · (−12)) =
(29) 29
0
v 1
v3 := p 03 0 = √ √ (3, 2, −12)
hv3 , v3 i 6 29
Thus we have the orthonormal basis

1 1 1
√ (1, 1, 1), √ √ (−7, 5, −1), √ √ (3, 2, −12) .
6 6 29 6 29
u = (1, 0, −1) and v = (3, 1, 0).
The Gram-Schmidt process yields
p 1
x := u/ hu, ui = √ (1, 0, −1)
2
3 1
y0 := v − hv, xix = (3, 1, 0) − (1, 0, −1) = (3, 2, 3)
2 2
p 1
y := y0 / hy0 , y0 i = √ (3, 2, 3)
22
104
MAT2611/101/3/2017
where we used the Euclidean inner product.

√ 16 1
w1 = hw, xix + hw, yiy = − 2x + √ y = (13, 16, 35)
22 11
1
w2 = w − w1 = (−2, 6, −2)
11
1 1
w= (13, 16, 35) + (−2, 6, −2) .
11 11
Question 29
(29.1) Let
 
0 0 2 0
1 0 1 0
A= 0 1 −2 0 .

0 0 0 1
The characteristic equation in λ for A is
   
1 0 0 0 λ 0 −2 0
 0 1 0 0  −1 λ −1 0
λ 0 0 1 0 − A = 0 −1 λ + 2
det    
0
0 0 0 1 0 0 0 λ−1
λ −1 0 0 −2 0
= λ −1 λ + 2 0 − (−1) −1 λ + 2 0
0 0 λ−1 0 0 λ−1
= λ(λ − 1)(λ(λ + 2) − 1) − 2(λ − 1)
= (λ − 1)(λ3 + 2λ2 − λ − 2)
= (λ − 1)(λ2 (λ + 2) − (λ + 2))
= (λ − 1)(λ + 2)(λ2 − 1)
= (λ − 1)2 (λ + 2)(λ + 1) = 0
where we used cofactor expansion along the first column. We find that the eigenvalues
of A are -2, -1 and 1 (twice).

             
1 0 0 0 0 0 2 0 a −2 0 −2 0 a −2(a + c) 0
 0
−2  1 0 0 1 0 1 0  b  −1 −2 −1 0   b  −a − 2b − c 0
            
− = =  = 0

 0 0 1 0 0 1 −2 0  c   0 −1 0 0  c  −b
0 0 0 1 0 0 0 1 d 0 0 0 −3 d −3d 0
Obviously b = d = 0 and c = −a. Thus the eigenvectors corresponding to the eigenvalue
-2 are   

 a 

  0  
  : a ∈ R \ {0} .
 −a 
 
0
 
105
The eigenspace is   

 a 

0
  
  : a∈R .
 −a 
 
0
 
A basis for this eigenspace is   


 1 
  
0 .

 −1
 
0
 

             
1 0 0 0 0 0 2 0 a −1 0 −2 0 a −a − 2c 0
 0
−  1 0 0 1 0 1 0  b  −1 −1 −1 0   b  −a − b − c 0
            
− = = = 
 0 0 1 0 0 1 −2 0  c   0 −1 1 0   c   −b + c  0
0 0 0 1 0 0 0 1 d 0 0 0 −2 d −2d 0
Obviously d = 0 and c = b = −a/2. Thus the eigenvectors corresponding to the

eigenvalue -1 are   

 a 

−a/2
  
  : a ∈ R \ {0} .
 −a/2 
 
0
 

 a 

−a/2
  
  : a∈R .
−a/2 
 
0
 
A basis for this eigenspace is  


 2 
 
−1 .

 −1
 
0
 

            
1 0 0 0 0 0 2 0 a 1 0 −2 0 a a − 2c 0
0 1 0 0 1 0 1 0  b  −1 1
       −1 0  b  −a + b − c 0
     
 − = = = 
0 0 1 0 0 1 −2 0  c   0 −1 3 0  c   −b + 3c  0
0 0 0 1 0 0 0 1 d 0 0 0 0 d 0 0
Obviously a = 2c and b = 3c. Thus the eigenvectors corresponding to the eigenvalue 1

are   

 2c 

3c
  
  : c ∈ R \ {0} .
 c 
 
d
 
106
MAT2611/101/3/2017

 2c 

3c
  
  : c, d ∈ R .
c 
 
d
 
A basis for this eigenspace is    


 2 0 

3 ,
  0
  .
 1 0
 
0 1
 
(29.2) Let T : P2 → P2 be the function defined by T (p(x)) = p(2x + 1).

We have for p(x) = p0 + p1 x + p2 x2 , q(x) = q0 + q1 x + q2 x2 ∈ P2 and k ∈ R:
1.
T ( (kp)(x) ) = T ((kp0 ) + (kp1 )x + (kp2 )x2 ) = (kp0 ) + (kp1 )(2x + 1) + (kp2 )(2x + 1)2
= k(p0 + p1 (2x + 1) + p2 (2x + 1)2 ) = k p(2x + 1) = kT (p(x))
2.
T ( (p + q)(x) ) = T ((p0 + q0 ) + (p1 + q1 )x + (p2 + q2 )x2 )

= (p0 + q0 ) + (p1 + q1 )(2x + 1) + (p2 + q2 )(2x + 1)2
= p0 + p1 (2x + 1) + p2 (2x + 1)2 + q0 + q1 (2x + 1) + q2 (2x + 1)2
= p(2x + 1) + q(2x + 1) = T (p(x)) + T (q(x))
so that T is linear.
(b) Find [T ]B with respect to the basis {1, x, x2 }.
From
T (1) = 1 · 1 + 0 · x + 0 · x2
T (x) = (2x + 1) = 1 · 1 + 2 · x + 0 · x2
T (x2 ) = (2x + 1)2 = 1 · 1 + 4 · x + 4 · x2
and the ordering

     
1 0 0
1 → 0 , x → 1 , x2 → 0
0 0 1
we find  
1 1 1
[T ]B = 0 2 4 .
0 0 4
107
(c) Compute T (2 − 3x + 4x2 ).
We find
T (2 − 3x + 4x2 ) = 2 − 3(2x + 1) + 4(2x + 1)2 = 2 − 6x − 3 + 16x2 + 16x + 4 = 3 + 10x + 16x2 .
or equivalently
  
2 3
T (2 − 3x + 4x2 ) = [T ]B −3 = 10 → 3 + 10x + 16x2 .
4 16
(29.3) Let S : P2 → P3 be defined by S(p(x)) = xp(x).

Let p(x) = p0 + p1 x + p2 x2 , q(x) = q0 + q1 x + q2 x2 ∈ P2 . Then S(p(x)) = p0 x +
p1 x2 + p2 x3 and S(q(x)) = q0 x + q1 x2 + q2 x3 so that
S(p(x)) − S(q(x)) = (p0 − q0 )x + (p1 − q1 )x2 + (p2 − q2 )x3 = 0 ⇔ p0 = q0 , p1 = q1 , p2 = q2

⇔ p(x) = q(x)
since x, x2 , and x3 are linearly independent.

(b) Find S −1 (p(x)).
Note that S −1 : {ax + bx2 + cx3 : a, b, c ∈ R} → P2 . The inverse is defined by
S −1 (S(p(x))) = p(x) ⇔ S −1 (xp(x)) = p(x)

S(S −1 (q(x))) = q(x) ⇔ xS −1 (q(x)) = q(x)
where p(x) ∈ P2 and q(x) ∈ {ax + bx2 + cx3 : a, b, c ∈ R}. Clearly
q(x)
S −1 (q(x)) = .
x
However, we consider this problem in terms of vector spaces where x is just a
placeholder. Let q(x) = q1 x + q2 x2 + q3 x3 and S −1 (q(x)) = s0 + s1 x + s2 x2 where
q1 , q2 , q3 , s0 , s1 , s2 ∈ R. Then
xS −1 (q(x)) = s0 x + s1 x2 + s2 x3 = q1 x + q2 x2 + q3 x3 = q(x)
so that s0 = q1 , s1 = q2 , s2 = q3 and
S −1 (q1 x + q2 x2 + q3 x3 ) = q1 + q2 x + q3 x2 .

No. Since R(S) = {ax+bx2 +cx3 : a, b, c ∈ R} and 1 ∈ P3 but 1 ∈
/ {ax+bx2 +cx3 :
a, b, c ∈ R}.
108
MAT2611/101/3/2017
Question 30
(30.1) Let
 
−2 0 0 0
 0 −2 5 −5 
A=
0
.
0 3 0 
0 0 0 3
Since A is upper triangular, the eigenvalues are the entries on the diagonal namely -2
(twice) and 3 (twice).

             
1 0 0 0 −2 0 0 0 a 0 0 0 0 a 0 0
 0
−2  1 0 0  0 −2 5 −5
    b  0 0 −5 5   b  −5c + 5d 0
 0 −    = 
  c  0 0 −5 0   c  =  −5c  = 0
     
0 1 0  0 0 3 0
0 0 0 1 0 0 0 3 d 0 0 0 −5 d −5d 0
Obviously c = d = 0. Thus the eigenvectors corresponding to the eigenvalue -2 are
  

 a 

b
  
2
  : (a, b) ∈ R \ {(0, 0)} .
 0 
 
0
 

 a 

b
  
  : a, b ∈ R .
 0 
 
0
 


 1 0 

0 ,
  1
  .
 0 0
 
0 0
 

             
1 0 0 0 −2 0 0 0 a 5 0 0 0 a 5a 0
 0 1 0 0  0 −2 5 −5   b  0 5 −5 5  b  5b − 5c + 5d 0
     
 0 0 1 0 −  0
3     =  =  = 0
  
0 3 0   c  0 0 0 0  c   0
0 0 0 1 0 0 0 3 d 0 0 0 0 d 0 0
Obviously a = 0 and b = c − d. Thus the eigenvectors corresponding to the eigenvalue
3 are   

 0 

 c − d 
  : (c, d) ∈ R2 \ {(0, 0)} .
  c  
 
d
 
109

 0 

c − d
  
  : c, d ∈ R .
  c  
 
d
 


 0 0 
−1
1 ,
 
  .
 1  0 
 
0 1
 
(30.2) Let T : P2 → P3 be the function defined by T (p(x)) = xp(x − 3).

We have for p(x) = p0 + p1 x + p2 x2 , q(x) = q0 + q1 x + q2 x2 ∈ P2 and k ∈ R:
1.
T ( (kp)(x) ) = T ((kp0 ) + (kp1 )x + (kp2 )x2 ) = x((kp0 ) + (kp1 )(x − 3) + (kp2 )(x − 3)2 )
= k(x(p0 + p1 (x − 3) + p2 (x − 3)2 )) = k xp(x − 3) = kT (p(x))
2.
T ( (p + q)(x) ) = T ((p0 + q0 ) + (p1 + q1 )x + (p2 + q2 )x2 )
= x((p0 + q0 ) + (p1 + q1 )(x − 3) + (p2 + q2 )(x − 3)2 )
= x(p0 + p1 (x − 3) + p2 (x − 3)2 ) + x(q0 + q1 (x − 3) + q2 (x − 3)2 )
= xp(x − 3) + xq(x − 3) = T (p(x)) + T (q(x))
(b) Find [T ]B 0 ,B with respect to the basis B = {1, x, x2 } and B 0 = {1, x, x2 , x3 }.
From
T (1) = x · [1]x→x−3 = x = 0 · 1 + 1 · x + 0 · x2 + 0 · x3
T (x) = x · [x]x→x−3 = x(x − 3) = 0 · 1 + (−3) · x + 1 · x2 + 0 · x3
T (x2 ) = x · [x2 ]x→x−3 = x(x − 3)2 = 0 · 1 + 9 · x + (−6) · x2 + 1 · x3
and the ordering
       
1 0 0 0
0 1 0 0
   
1→
0 ,
 x→
0 ,
 x2 →   3
1 x → 0
0 0 0 1
we find (the columns are the coefficients of the elements of B 0 in T (1), T (x) and
T (x2 ) above)  
0 0 0
1 −3 9 
[T ]B 0 ,B = 
0 1 −6 .

0 0 1
110
MAT2611/101/3/2017
(c) Compute T (1 + x − x2 ).
We find
T (1 + x − x2 ) = x(1 + (x − 3) − (x − 3)2 ) = −11x + 7x2 − x3 .
or equivalently
 
  0
1 −11
2 2 3
T (1 + x − x ) = [T ]B 0 ,B 1  = 

 7  → −11x + 7x − x .

−1
−1
(30.3) Let S : P3 → P3 be defined by S(p(x)) = p(x + 1).

Let p(x) = p0 +p1 x+p2 x2 +p3 x3 , q(x) = q0 +q1 x+q2 x2 +q3 x3 ∈ P2 . Then S(p(x)) =
p0 +p1 (x+1)+p2 (x+1)2 +p3 (x+1)3 and S(q(x)) = q0 +q1 (x+1)+q2 (x+1)2 +q3 (x+1)3
so that
S(p(x)) − S(q(x)) = (p0 − q0 ) + (p1 − q1 )(x + 1) + (p2 − q2 )(x + 1)2 + (p3 − q3 )(x + 1)3 = 0
= ((p0 − q0 ) + (p1 − q1 ) + (p2 − q2 ) + (p3 − q3 ))
+ ((p1 − q1 ) + 2(p2 − q2 ) + 3(p3 − q3 ))x
+ ((p2 − q2 ) + 3(p3 − q3 ))x2
+ (p3 − q3 )x3
Setting each coefficient (from x3 down to x0 ) to zero we find p3 −q3 = 0, p2 − q2 = 0,

p1 − q1 = 0 and p0 − q0 = 0 so that p(x) = q(x) since 1, x, x2 , and x3 are linearly
independent.
(b) Find S −1 (p(x)).
The inverse is defined by
S −1 (S(p(x))) = p(x) ⇔ S −1 (p(x + 1)) = p(x)

S(S −1 (q(x))) = q(x) ⇔ S −1 (q(x)) = q(x)
x→x+1
where p(x), q(x) ∈ P3 . Clearly
S −1 (q(x)) = q(x − 1).
However, we consider this problem in terms of vector spaces where x is just a

placeholder. Let q(x) = q0 +q1 x+q2 x2 +q3 x3 and S −1 (q(x)) = s0 +s1 x+s2 x2 +s3 x3
where q0 , q1 , q2 , q3 , s0 , s1 , s2 , s3 ∈ R. Then
S(S −1 (q(x))) = s0 + s1 (x + 1) + s2 (x + 1)2 + s3 (x + 1)3 = q0 + q1 x + q2 x2 + q3 x3 = q(x)

⇔ s0 + s1 + s2 + s3 = q0 , s1 + 2s2 + 3s3 = q1 , s2 + 3s3 = q2 , s3 = q3 (?)
111
so that s3 = q3 , s2 = q2 − 3q3 , s1 = q1 − 2q2 + 3q3 and s0 = q0 − q1 + q2 − q3 and
S −1 (q0 + q1 x + q2 x2 + q3 x3 ) = (q0 − q1 + q2 − q3 ) + (q1 − 2q2 + 3q3 )x + (q2 − 3q3 )x2 + q3 x2

= q0 + q1 (x − 1) + q2 (x − 1)2 + q3 (x − 1)3 = q(x − 1).

Yes. Since from (??) for every q(x) ∈ P3 there exists s0 + s1 x + s2 x2 + s3 x3 ∈ P3
(where s0 , s1 , s2 , s3 ∈ R) such that S(s0 + s1 x + s2 x2 + s3 x3 ) = q(x).
Alternatively, since S is one-to-one and the kernel of S is {0}, we find that S is

onto.
Question 31
T (a0 + a1 x + a2 x2 ) = (2a0 − a1 + 3a2 ) + (4a0 − 5a1 )x + (a1 + 2a2 )x2 .

We determine the matrix representation for T with respect to the basis B = {1, x, x2 }
in that order. We have
T (1) = 2 · 1 + 4 · x + 0 · x2
T (x) = −1 · 1 − 5 · x + 1 · x2
T (x2 ) = 3 · 1 + 0 · x + 2 · x2
and the ordering      

1 0 0
1 → 0 , x → 1 , x2 → 0
0 0 1
so that  
2 −1 3
[T ]B = 4 −5 0 .
0 1 2
  
λ 0 0 λ−2 1 −3
det  0 λ 0 − [T ]B = −4 λ + 5
  0
0 0 λ 0 −1 λ − 2
= (λ + 5)(λ − 2)2 − 12 + 4(λ − 2) = λ3 + λ2 − 12λ = λ(λ + 4)(λ − 3).
Thus the eigenvalues are -4, 0 and 3.
112
MAT2611/101/3/2017

The eigenvectors of T have matrix representations given by the eigenvectors of [T ]B .
We determine the eigenvectors for the eigenvalue −4 of [T ]B :
             
1 0 0 2 −1 3 x −6 1 −3 x −6x + y − 3z 0
−4 0 1 0 − 4 −5 0 y  = −4 1 0   y =
  −4x + y  = 0

0 0 1 0 1 2 z 0 −1 −6 z −y − 6z 0
Obviously 4x = y = −6z. Thus the eigenvectors corresponding to the eigenvalue -4 are

  
 −3z/2 
 −6z  : z ∈ R \ {0} .
z
 
Setting z = 2, for example, yields the basis
{ −3 − 12x + 2x2 }.
We determine the eigenvectors for the eigenvalue 0 of [T ]B :

             
1 0 0 2 −1 3 x −2 1 −3 x −2x + y − 3z 0
0 0 1 0 − 4 −5 0 y  = −4 5 0  y  =  −4x + 5y  = 0
0 0 1 0 1 2 z 0 −1 −2 z −y − 2z 0
Obviously y = −2z, x = 5y/4 = −5z/2. Thus the eigenvectors corresponding to the

eigenvalue 0 are   
 −5z/2 
 −2z  : z ∈ R \ {0} .
z
 
{ −5 − 4x + 2x2 }.

             
1 0 0 2 −1 3 x 1 1 −3 x x + y − 3z 0
3 0 1 0 − 4 −5 0 y  = −4 8 0   y = −4x + 8y = 0
   
0 0 1 0 1 2 z 0 −1 1 z −y + z 0
Obviously y = z, x = 2y = 2z. Thus the eigenvectors corresponding to the eigenvalue 3

are   
 2z 
 z  : z ∈ R \ {0} .
z
 
{ 2 + x + x2 }.
113
Question 32
T (a0 + a1 x + a2 x2 ) = −2a2 + (a0 + 2a1 + a2 )x + (a0 + 3a2 )x2 .

We determine the matrix representation for T with respect to the basis B = {1, x, x2 }
in that order. We have
(a0 = 1, a1 = 0, a2 = 0) T (1) = x + x2 = 0 · 1 + 1 · x + 1 · x2
(a0 = 0, a1 = 1, a2 = 0) T (x) = 2x = 0 · 1 + 2 · x + 0 · x2
(a0 = 0, a1 = 0, a2 = 1) T (x2 ) = −2 + x + 3x2 = −2 · 1 + 1 · x + 3 · x2
and the ordering      

1 0 0
1 → 0 , x → 1 , x2 → 0
0 0 1
so that  
0 0 −2
[T ]B = 1 2 1  .
1 0 3
  
λ 0 0 λ 0 2
det   0 λ 0 − [T ]B = −1 λ − 2 −1
 
0 0 λ −1 0 λ−3
= λ(λ − 2)(λ − 3) + 2(λ − 2)
= (λ(λ − 3) + 2)(λ − 2) = (λ − 1)(λ − 2)2 .
Thus the eigenvalues are 1 and 2 (twice).

The eigenvectors of T have matrix representations given by the eigenvectors of [T ]B .
             
1 0 0 0 0 −2 a 1 0 2 a a + 2c 0
1 0 1 0 − 1 2 1   b  = −1 −1 −1  b  = −a − b − c = 0
0 0 1 1 0 3 c −1 0 −2 c −a − 2c 0
Obviously a = −2b = −2c. Thus the eigenvectors corresponding to the eigenvalue 1 are
  
 −2c 
 c  : c ∈ R \ {0} .
c
 
114
MAT2611/101/3/2017
Setting c = 1, for example, yields the basis
{ −2 + x + x2 }.

             
1 0 0 0 0 −2 a 2 0 2 a 2a + 2c 0
2 0 1 0 − 1 2 1
     b = −1 0 −1
    b = −a − c = 0
   
0 0 1 1 0 3 c −1 0 −1 c −a − c 0
Obviously a = −c. Thus the eigenvectors corresponding to the eigenvalue 2 are

  
 a 
 b  : b ∈ R \ {0} .
−a
 
Setting a = 0, b = 1, and then a = 1, b = 0 for example, yields the basis
{ x, 1 − x2 }.
Question 33

   
x −x + 2y + 4z
T y  =  3x + z .
z 2x + 2y + 5z
basis B for R3 .
A set of 3 linearly independent eigenvectors of T will diagonalize T . We find the repre-
sentation of the eigenvectors in the standard basis. For the standard basis we have
       
1 1 0 0
T 0 = −1 0 + 3 1 + 2 0
0 0 0 1
        
0 1 0 0
T 1 = 2 0 + 0 1 + 2 0
0 0 0 1
       
0 1 0 0
T   0  = 4 0 + 1 1 + 5 0
    
1 0 0 1
so that (with the usual ordering of the standard basis)

 
−1 2 4
[T ]B = 3 0
 1 .
2 2 5
115
First we need the eigenvalues an eigenvectors of [T ]B . The characteristic equation is
  
λ 0 0 λ+1 −2 −4
det   0 λ 0 − [T ]B = −3
  λ −1
0 0 λ −2 −2 λ − 5
= λ3 − 4λ2 − 21λ = λ(λ − 7)(λ + 3).
Thus the eigenvalues are 0, -3 and 7 . We determine the eigenvectors for the eigenvalue
0:
             
1 0 0 −1 2 4 x 1 −2 −4 x x − 2y − 4z 0
0 0 1 0 −  3 0 1 y  = −3 0 −1 y  =  −3x − z  = 0
0 0 1 2 2 5 z −2 −2 −5 z −2x − 2y − 5z 0
Obviously z = −3x and y = (x − 4z)/2 = 13x/2. Thus the eigenvectors corresponding

to the eigenvalue 0 are   
 x 
13x/2 : x ∈ R \ {0} .
−3x
 
A representative eigenvector is (any one could have been chosen)

 
2
 13  .
−6

             
1 0 0 −1 2 4 x −2 −2 −4 x −2x − 2y − 4z 0
−3 0 1 0 −  3 0 1 y  = −3 −3 −1  y = −3x − 3y − z = 0
   
0 0 1 2 2 5 z −2 −2 −8 z −2x − 2y − 8z 0
We have −2x−2y −4z = −2x−2y −8z = 0 so that z = 0. It follows that −3x−3y −z =

−3x − 3y = 0 and x = −y. Thus the eigenvectors corresponding to the eigenvalue -3
are   
 x 
−x : x ∈ R \ {0} .
0
 

 
1
−1 .
0

             
1 0 0 −1 2 4 x 8 −2 −4 x 8x − 2y − 4z 0
7 0 1 0 −  3 0 1 y  = −3 7 −1  y = −3x + 7y − z = 0
   
0 0 1 2 2 5 z −2 −2 2 z −2x − 2y + 2z 0
116
MAT2611/101/3/2017
Row reduction yields

   
+4R3 8 −2 −4 0 =R3 0 −10 4 0
−3R3 /2  −3 7 −1 0  → +R1  0 10 −4 0 
−R3 /2 −2 −2 2 0 =R2 1 1 −1 0
 
1 −1 1 0
→  0 0 0 0.
0 10 −4 0
Obviously y = 2z/5 and x = −y + z = 3z/5. Thus the eigenvectors corresponding to

the eigenvalue 7 are   
 3z/5 
2z/5 : z ∈ R \ {0} .
z
 

 
3
2 .
5
Thus a basis B 0 relative to which the matrix for T is diagonal is

     
 2 1 3 
0
B =  13 , −1 ,
  2 .
−6 0 5
 
(33.2) Compute [T ]B 0 and verify that [T ]B 0 = P −1 [T ]B P where the matrix P diagonalizes [T ]B .

For the basis B 0 above we have
          
2 0 2 1 3
T   13   = 0 = 0 13 + 0 −1 + 0 2
      
−6 0 −6 0 5
          
1 −3 2 1 3
T  −1   = 3 = 0 13 − 3 −1 + 0 2
      
0 0 −6 0 5
         
3 21 2 1 3
T   2  = 14 = 0 13 + 0 −1 + 7 2
      
5 35 −6 0 5
so that (with the ordering in the set B 0 above)

 
0 0 0
[T ]B 0 = 0 −3 0 .
0 0 7
117
 
2 1 3
P =  13 −1 2 .
−6 0 5
matrix
   
=R1 /2 2 1 3 1 0 0 1 1/2 3/2 1/2 0 0
−13R1 /2  13 −1 2 0 1 0  → +5R3 /2  0 −15/2 −35/2 −13/2 1 0 
+3R1 −6 0 5 0 0 1 0 3 14 3 0 1
 
1 1/2 3/2 1/2 0 0
→ =2R2 /35  0 0 35/2 1 1 5/2 
0 3 14 3 0 1
 
−3R2 /2 1 1/2 3/2 1/2 0 0
→  0 0 1 2/35 2/35 1/7 
−14R2 0 3 14 3 0 1
 
−R3 /6−3R2 /2 1 1/2 3/2 1/2 0 0
→ =R3 /3  0 0 1 2/35 2/35 1/7 
=R2 /3 0 3 0 11/5 −4/5 −1
 
1 0 0 1/21 1/21 −1/21
→  0 1 0 11/15 −4/15 −1/3  .
0 0 1 2/35 2/35 1/7
Thus  
1/21 1/21 −1/21
1
P −1 = 11/15 −4/15 −1/3  .
2
2/35 2/35 1/7
It follows that
   
1/21 1/21 −1/21 −1 2 4 2 1 3
−1
P [T ]B P = 11/15
 −4/15 −1/3   3 0 1  13 −1 2
2/35 2/35 1/7 2 2 5 −6 0 5
 
0 0 0
= 0 −3
 0 = [T ]B 0 .
0 0 7
Question 34

   
x 4x + z
T y  = 2x + 3y + 2z  .
z x + 4z
118
MAT2611/101/3/2017
basis B for R3 .
A set of 3 linearly independent eigenvectors of T will diagonalize T . We find the repre-
sentation of the eigenvectors in the standard basis. For the standard basis we have
       
1 1 0 0
T   0 = 4 0 + 2 1 + 0
    
0 0 0 1
       
0 1 0 0
T   1 = 0 0 + 3 1 + 0 0
    
0 0 0 1
       
0 1 0 0
T   0 = 0 + 2 1 + 4 0
    
1 0 0 1
so that (with the usual ordering of the standard basis)
 
4 0 1
[T ]B = 2 3 2 .
1 0 4
First we need the eigenvalues an eigenvectors of [T ]B . The characteristic equation is
  
λ 0 0 λ−4 0 −1
det   0 λ 0 − [T ]B = −2 λ − 3 −2
 
0 0 λ −1 0 λ−4
= (λ − 3)(λ − 4)2 − (λ − 3) = (λ − 3)((λ − 4)2 − 1) = (λ − 3)2 (λ − 5).
Thus the eigenvalues are 3 (twice) and 5 . We determine the eigenvectors for the
eigenvalue 3:
             
1 0 0 4 0 1 x −1 0 −1 x −x − z 0
3 0 1 0 − 2 3 2 y  = −2 0 −2  y = −2x − 2z = 0
   
0 0 1 1 0 4 z −1 0 −1 z −x − z 0
Obviously x = −z. Thus the eigenvectors corresponding to the eigenvalue 3 are
  
 x 
 y  : (x, y) ∈ R2 \ {(0, 0)} .
−x
 
Two representative eigenvectors are (any two linearly independent eigenvectors could
have been chosen)    
1 0
 0 , 1 .
−1 0
             
1 0 0 4 0 1 x 1 0 −1 x x−z 0
5 0 1 0 − 2 3 2 y  = −2 2 −2 y  = −2x + 2y − 2z  = 0
0 0 1 1 0 4 z −1 0 1 z −x + z 0
119
Obviously x = z and y = 2x. Thus the eigenvectors corresponding to the eigenvalue 5
are   
 x 
2x : x ∈ R \ {0} .
x
 

 
1
2 .
1
Thus a basis B 0 relative to which the matrix for T is diagonal is
     
 1 0 1 
B 0 =  0  , 1 , 2 .
−1 0 1
 
(34.2) Compute [T ]B 0 and verify that [T ]B 0 = P −1 [T ]B P where the matrix P diagonalizes [T ]B .

For the basis B 0 above we have
         
1 3 1 0 1
T  0  =  0  = 3  0  + 0 1 + 0 2
−1 −3 −1 0 1
         
0 0 1 0 1
T 1 = 3 = 0  0  + 3 1 + 0 2
0 0 −1 0 1
         
1 5 1 0 1
T 2 = 10 = 0  0  + 0 1 + 5 2
1 5 −1 0 1
so that (with the ordering in the set B 0 above)
 
3 0 0
[T ]B 0 = 0 3 0 .
0 0 5
 
1 0 1
P =  0 1 2 .
−1 0 1
matrix
   
1 0 1 1 0 0 −R3/2 1 0 1 1 0 0
 0 1 2 0 1 0 → − R3  0 1 2 0 1 0 
+R1 −1 0 1 0 0 1 =R3/2 0 0 2 1 0 1
1 0 0 12 0 − 12
 
→  0 1 0 −1 1 −1  .
0 0 1 12 0 21
120
MAT2611/101/3/2017
Thus  
1 0 −1
1
P −1 = −2 2 −2 .
2
1 0 1
It follows that
   
1 0 −1 4 0 1 1 0 1
1
P −1 [T ]B P = −2 2 −2 2 3 2  0 1 2
2
1 0 1 1 0 4 −1 0 1
  
1 0 −1 3 0 5
1
= −2 2 −2  0 3 10
2
1 0 1 −3 0 5
 
3 0 0
= 0 3
 0 = [T ]B 0 .
0 0 5
Question 35
 
−1 2 4
Let T : R3 → R3 be multiplication by the matrix A =  3 0 1 . Find
2 2 5

The range of T is the column space of A. Applying row reduction to AT we find
       
−1 3 2 −1 3 2 −1 3 2 −1 3 2
+2 R1 2 0 2 → =R2/6 0 6 6  →  0 1 1 →  0 1 1 .
+4 R1 4 1 5 =R3/13 0 13 13 −R2 0 1 1 0 0 0
Thus a basis for the range of T is

   
 −1 0 
 3 , 1 .
2 1
 

The kernel of T is the nullspace of A. Let x, y, z ∈ R satisfy
      
x −1 2 4 x 0
A  y  =  3 0 1   y  = 0 .

z 2 2 5 z 0

     
−1 2 4 −1 2 4 −1 2 4
+3 R1 3 0 1 →  0 6 13 → =R2/6 0 6 13 .
+2 R1 2 2 5 −R2 0 6 13 =R3/13 0 0 0
121
Thus we have the nullpsace of A y = −13z/6 and x = 2y + 4z = −z/3.
   
 z −2 
−13 , z ∈ R
6
6

which has a basis  

 −2 
−13 .
6
 

From the basis found in 35.1 we find the rank of T is 2.
From the basis found in 35.2 we find the nullity of T is 1.

From the row reduction in 35.2 we find the rank of A is 2.
From 35.2 we find the nullity of A is 1.
Question 36
 
1 0 3
Let T : R3 → R3 be multiplication by the matrix A = 1 2 4 . Find
1 8 5

The range of T is the column space of A. Applying row reduction to AT we find
     
1 1 1 1 1 1 1 1 1
0 2 8 → = R2 /2 0 2 8 → 0 1 4  .
−3 R1 3 4 5 − R2 /2 0 1 2 0 0 −2
Thus a basis for the range of T is

     
 1 0 0 
1 , 1 ,  0  .
1 4 −2
 
In fact, any basis for R3 will do, including the standard basis (or the same three columns),
since the three columns are linearly independent.
122
MAT2611/101/3/2017

Since A is invertible, the kernel consists only of the zero vector.
Alternative:
The kernel of T is the nullspace of A. Let x, y, z ∈ R satisfy
      
x 1 0 3 x 0
A y = 1 2 4
     y = 0 .
 
z 1 8 5 z 0

     
1 0 3 1 0 3 1 0 3
−R1 1 2 4 → 0 2 1 → 0 2 1  .
−R1 1 8 5 −4R1 0 8 2 0 0 −2
z = 0, y = −z/2 = 0 and x = −3z = 0. Thus we have the nullpsace of A

 
 0 
0
0
 
which has no basis.

From the basis found in 36.1 we find the rank of T is 3.
From the basis found in 36.2 we find the nullity of T is 0.

From the row reduction in 36.2 we find the rank of A is 3.
From 36.2 we find the nullity of A is 0.
Question 37

1 2a b
2
Consider T : P2 → M22 given by T (a + bx + cx ) = for all a, b, c ∈ R.
2 b 2c
• For all a, b, c, α, β, γ ∈ R
T ((a + bx + cx2 ) + (α + βx + γx2 )) = T ((a + α) + (b + β)x + (c + γ)x2 )

1 2(a + α) b + β 1 2a + 2α b + β
= =
2 b+β 2(c + γ) 2 b + β 2c + 2γ

2 2 1 2a b 1 2α β 1 2a + 2α b + β
T (a + bx + cx ) + T (α + βx + γx ) = + =
2 b 2c 2 β 2γ 2 b + β 2c + 2γ
i.e. T ((a + bx + cx2 ) + (α + βx + γx2 )) = T (a + bx + cx2 ) + T (α + βx + γx2 ).
123
• For all a, b, c, k ∈ R
T (k(a + bx + cx2 )) = T ((ka) + (kb)x + (kc)x2 )

1 2(ka) kb 1 2a b
= =k = kT (a + bx + cx2 ).
2 kb 2(kc) 2 b 2c
Thus T is a linear transform.

(37.2) Find the matrix representation for T relative to the standard basis in P2 and in M22
Applying T to the elements of the standard basis BP2 of P2 yields

2 1 2 0 1 0 0 1 0 0 0 0
T (1) = T (1 + 0x + 0x ) = =1 +0 +0 +0
2 0 0 0 0 0 0 1 0 0 1

2 1 0 1 1 0 1 0 1 1 0 0 0 0
T (x) = T (0 + 1x + 0x ) = =0 + + +0
2 1 0 0 0 2 0 0 2 1 0 0 1

2 2 1 0 0 1 0 0 1 0 0 0 0
T (x ) = T (0 + 0x + 1x ) = =0 +0 +0 +1 .
2 0 2 0 0 0 0 1 0 0 1
The coefficients of the elements of the standard basis BM22 in M22 provide the columns
of the matrix representation:
 
1 0 0
0 1/2 0
[T ]BP2 ,BM22 = 
0 1/2 0 .

0 0 1
(37.3) Is T invertible?
We have

2 1 2a b
R(T ) = T (a + bx + cx ) : a, b, c ∈ R = : a, b, c ∈ R .
2 b 2c
If we can solve for T −1 : R(T ) → P2 satisfying
(T −1 ◦ T )(a + bx + cx2 ) = T −1 (T (a + bx + cx2 )) = a + bx + cx2 ,

(TI)

−1 1 2a b −1 1 2a b 1 2a b
(T ◦ T ) =T T = .
2 b 2c 2 b 2c 2 b 2c
for all a, b, c ∈ R then T is invertible. Suppose

−1 1 2a b
T = α + βx + γx2
2 b 2c
then

−1 1 2a b 2 1 2α β 1 2a b
(T ◦ T ) = T (α + βx + γx ) = = .
2 b 2c 2 β 2γ 2 b 2c
124
MAT2611/101/3/2017
Clearly α = a, β = b and γ = c. Thus

−1 1 2a b
T = a + bx + cx2 .
2 b 2c
Also,

−1 2 −1 2 −1 1 2a b
(T ◦ T )(a + bx + cx ) = T (T (a + bx + cx )) = T = a + bx + cx2 .
2 b 2c
Since both equations (TI) are satisfied, T is invertible.
(37.4) Show that the range of T is the subspace M

f22 of M22 consisting of symmetric matrices.
We have
T (a + bx + cx2 ) : a, b, c ∈ R

R(T ) =

a b/2
= : a, b, c ∈ R
b/2 c
a b0

0
= : a, b , c ∈ R (set b0 = b/2)
b0 c
=Mf22 .
(37.5) Let Te : P2 → M
f22 be defined by Te(p(x)) := T (p(x)) for all p(x) ∈ P2 . Find the matrix
representation for Te relative to the standard basis with the usual ordering in P2 and the
basis
1 0 0 0 0 1
, ,
0 0 0 1 1 0
Applying Te to the elements of the standard basis BP2 of P2 yields

2 1 2 0 1 0 0 0 0 1
T (1) = T (1 + 0x + 0x ) =
e e =1 +0 +0
2 0 0 0 0 0 1 1 0

1 0 1 1 0 0 0 1 0 1
Te(x) = Te(0 + 1x + 0x2 ) = =0 +0 +
2 1 0 0 0 0 1 2 1 0

2 2 1 0 0 1 0 0 0 0 1
Te(x ) = Te(0 + 0x + 1x ) = =0 +1 +0 .
2 0 2 0 0 0 1 1 0
The coefficients of the elements of the basis BM

f22 in M22 provide the columns of the
f
matrix representation:  
1 0 0
[Te]BP2 ,BMf = 0 0 1 .
22
0 1/2 0
125
Question 38
Consider T : M22 → P2 given by

1
T (A) = 1 x A , for all A ∈ M22 .
x
• By distributivity of the matrix product, we have for all A, B ∈ M22

1
T (A + B) = 1 x (A + B)
x

1
= 1 x A+ 1 x B
x

1 1
= 1 x A + 1 x B
x x
= T (A) + T (B).
• For all k ∈ R and A ∈ M22 we have

1 1
T (A) = 1 x kA =k 1 x A = kT (A).
x x
Thus T is a linear transform.

(38.2) Find the matrix representation for T relative to the standard basis in M22 and in P2
Applying T to the elements of the standard basis BM22 of M22 yields

1 0
T = 1 + 0x + 0x2
0 0

0 1
T = 0 + 1x + 0x2
0 0

0 0
T = 0 + 1x + 0x2
1 0

0 0
T = 0 + 0x + 1x2 .
0 1
The coefficients of the elements of the standard basis BP2 in P2 provide the columns of
the matrix representation:
 
1 0 0 0
[T ]BM22 ,BP2 = 0 1 1 0 .
0 0 0 1
126
MAT2611/101/3/2017
(38.3) Is T one to one?

Since
0 1
T = 2x
1 0
and
0 2
T = 2x
0 0
T is not one to one.
(38.4) Let Mf22 be the subspace of M22 consisting of symmetric matrices. Let Te : Mf22 → P2 be
defined by Te(A) := T (A) for all A ∈ M
f22 . Find the matrix representation for Te relative
to the standard basis with the usual ordering in P2 and the basis

1 0 0 0 0 1
, ,
0 0 0 1 1 0

Applying T to the elements of the given basis B of M f22

1 0
T = 1 + 0x + 0x2
0 0

0 0
T = 0 + 0x + 1x2
0 1

0 1
T = 0 + 2x + 0x2 .
1 0
The coefficients of the elements of the standard basis BP2 in P2 provide the columns of
the matrix representation:  
1 0 0
[T ]B,BP2 = 0 0 2 .
0 1 0
Question 39
Consider T : P2 → P2 given by T (a + bx + cx2 ) = b + cx + ax2 for all a, b, c ∈ R.
(39.1)
(a) Find the kernel and nullity of T .
The kernel of T is given by
ker(T ) = { a + bx + cx2 : a, b, c ∈ R, T (a + bx + cx2 ) = 0 + 0x + 0x2 }

= { a + bx + cx2 : a, b, c ∈ R, b + cx + ax2 = 0 + 0x + 0x2 }
= { a + bx + cx2 : a = 0, b = 0, c = 0 } = { 0 + 0x + 0x2 }.
127
The nullity of T is the dimension of ker(T ) which is 0.
(b) Find the range and rank of T .
The range of T is given by
R(T ) = { T (a + bx + cx2 ) : a, b, c ∈ R }
= { b + cx + ax2 : a, b, c ∈ R } = P2 .
The rank of T is the dimension of R(T ) which is 3.
(39.2) Find the real valued eigenvalues and corresponding eigenspaces of T .

The eigenvalue equation for T is
T (a + bx + cx2 ) = λ(a + bx + cx2 ) ⇒ b + cx + ax2 = (λa) + (λb)x + (λc)x2
where a, b, c, λ ∈ R, (a, b, c) 6= (0, 0, 0). Comparing coefficients of elements of the stan-

dard basis in P2 yields
b = λa, c = λb = λ2 a, a = λc = λ3 a.
If a = 0, then b = 0 and c = 0 which contradicts (a, b, c) 6= (0, 0, 0). So a 6= 0.

Consequently λ3 = 1 which has the only real solution λ = 1. From λ = 1 we find
b = λa = a and c = λb = b = a. Thus 1 is the only real eigenvalue of T and the
corresponding eigenspace is
{ a + ax + ax2 : a ∈ R } = { a(1 + x + x2 ) : a ∈ R }.
Alternative:
We find the matrix representation of T with respect to the standard basis B = {1, x, x2 }
in P2 , with the usual left to right ordering. We have
T (1) = T (1 + 0x + 0x2 ) = 0 + 0x + 1x2

T (x) = T (0 + 1x + 0x2 ) = 1 + 0x + 0x2
T (x2 ) = T (0 + 0x + 1x2 ) = 0 + 1x + 0x2
so that  
0 1 0
[T ]B = 0 0 1
1 0 0
The characteristic equation in λ for [T ]B is
    
1 0 0 0 1 0 λ −1 0
det λ 0 1 0 − 0 0 1 = 0 λ −1 = λ3 − 1 = 0.
0 0 1 1 0 0 −1 0 λ
The only real solution is λ = 1. Solving

           
1 0 0 0 1 0 a 1 −1 0 a 0
1 0 1 0 − 0 0 1    b = 0
  1 −1   b = 0
 
0 0 1 1 0 0 c −1 0 1 c 0
128
MAT2611/101/3/2017
for a, b, c ∈ R yields a = b, b = c and a = c so that a = b = c. The eigenspace of T

corresponding to the eigenvalue 1 is given by
   
 a 
p(x) ∈ P2 : [p(x)]B = a , a ∈ R = p(x) ∈ P2 : p(x) = a · 1 + a · x + a · x2 , a ∈ R
a
 
= a(1 + x + x2 ) : a ∈ R .

(39.3) Find T 3 := T ◦ T ◦ T .
We have
T 3 (a + bx + cx2 ) = T (T (T (a + bx + cx2 )))

= T (T (b + cx + ax2 ))
= T (c + ax + bx2 )
= a + bx + cx2
for all a, b, c ∈ R, i.e. T 3 is the identity on P2 and T 2 = T −1 .
129
F.2 Previous multiple choice questions
Questions
Question 1
Consider the set

X := { (x, y) : x, y ∈ R }
and the operations (for all k, x, y, α, β ∈ R, a = (x, y) ∈ X and b = (α, β) ∈ X)
· : R × X → X, k · a ≡ k · (x, y) := (kx − k + 1, ky),

+ : X × X → X, a + b ≡ (x, y) + (α, β) := (x + α − 1, y + β).
1. (1, 0)
2. (1, 1)
3. (0, 1)
4. (0, 0)
Question 2
Consider the set

X := { (x, y) : x, y ∈ R }
· : R × X → X, k · a ≡ k · (x, y) := (kx − k + 1, ky),

+ : X × X → X, a + b ≡ (x, y) + (α, β) := (x + α − 1, y + β).
1. −(1, 0) = (−1, 0)
2. −(1, 0) = (1, 1)
3. −(1, 0) = (1, 0)
4. −(1, 0) = (0, 0)
130
MAT2611/101/3/2017
Question 3
Which of the following are subspaces of R2 with the usual operations ?
A. span { (2, 3) }
B. { (1, x) : x ∈ R }
C. { (0, x) : x ∈ R, x ≥ 0 }
D. { (x, x) : x ∈ R }
1. Only A.
2. Only A and D.
3. Only C.
4. Only C and D.
Question 4
A. span { (π, 0) }
B. { (2, x) : x ∈ R }
C. { (x, y) : x, y ∈ N }
D. { (x, −x) : x ∈ R }
1. Only A and D.
2. Only A, B and D.
3. Only C.
4. Only D.
131
Question 5
A. span { (2, 3) } in R2
B. { (1, 1), (1, −1) } in R2
C. { (1, 1), (1, −1), (0, 1) } in R2
D. { 1 + x, 1 − x } in P1
1. Only A.
2. Only B.
3. Only B and C.
4. Only B and D.
Question 6
A. span { (π, 0) } in R2
B. { (1, 2), (2, 1) } in R2

0 1 0 1 0 0
C. , , in M22
1 0 −1 0 1 0
D. { 1 + 2x, 2 − x } in P1
1. Only B and D.
2. Only B.
3. Only C and D.
4. Only D.
132
MAT2611/101/3/2017
Question 7
Which of the following sets are identical?
A. span { (1, 0, 1), (1, 0, −1) } in R3
B. span { (0, 0, 1), (5, 0, 0) } in R3
C. span { (3, 0, 7), (0, 0, 1), (5, 0, 0) } in R3
D. { (1, 0, 1), (1, 0, −1) }
E. span { (1, 1, 1), (1, −1, −1) } in R3
1. Only A and D.
2. Only A and E.
3. Only A, B and C.
4. Only B and C.
Question 8
A. span { (1, 0, 1), (2, 0, 2) } in R3
B. span { (1, 0, 1), (1, 0, −1) } in R3
C. span { (1, 0, 1), (1, 1, 1) } in R3
D. span { (3, 0, 3) } in R3
E. { (1, 0, 1), (2, 0, 2) }
1. Only A and B.
2. Only A and D.
3. Only A and E.
4. Only B and C.
133
Question 9
Which of the following sets are a basis for the following vector subspace of P2 :
X = { p(x) ∈ P2 : p(3) = 0 } .
A. { 1, x, x2 }
B. { x − 3, x2 − 9 }
C. { x2 + 2x − 15, x2 − 2x − 3 }
D. { x − 3, x3 − 27 }
1. Only B and C.
2. Only B.
3. Only A.
4. Only B and D.
Question 10

1 0
X = A ∈ M22 : A = .
2 0

2 0 0 −1 0 0 0 0
A. , , ,
0 0 0 0 2 0 0 −1

2 −1
B.
2 −1

2 −1 0 0
C. ,
0 0 2 −1

2 −1 2 −1
D. ,
2 −1 −2 1
1. Only A.
2. Only B.
3. Only C.
4. Only C and D.
134
MAT2611/101/3/2017
Question 11
A. dim(span { (1, 0, 1), (1, 0, −1) }) = 2 in R3
B. dim(span { (0, 0, 0), (5, 0, 0) }) = 2 in R3
C. dim(span { (3, 0, 7), (0, 0, 1), (5, 0, 0) }) = 2 in R3
D. dim(span { (3, 0, 7), (0, 0, 1), (5, 0, 0) }) = 3 in R3
2. Only A and B.
3. Only A and C.
4. Only A and D.
Question 12
A. dim(span { (1, 0, 1), (1, 0, −1) }) = 2 in R3
B. dim(span { (0, 0, 0), (5, 0, 0) }) = 2 in R3
C. dim(span { (1, 0, 1), (1, 0, −1), (5, 0, 0) }) = 2 in R3
D. dim(span { (1, 0, 1), (1, 0, −1), (5, 0, 0) }) = 3 in R3
2. Only A, B and D.
3. Only A and C.
4. Only A and D.
135
Question 13

1 −1 2
Which of the following sets are a basis for the row space of ?
1 2 −1

A. 0 −3 3 , 1 2 −1

B. −1 1 1 , 0 0 1

C. 1 −1 2 , 1 2 −1
1. Only A.
2. Only C.
3. Only A and C.
4. Only B.
Question 14
 
2 −4
Which of the following sets are a basis for the row space of −1 2  ?
2 −4
n T T o
A. 2 −1 2 , −4 2 −4

B. 2 −4 , −1 2 , 2 −4

C. 2 −4 , −1 2

D. 1 2

E. 2 −1
1. Only A.
2. Only B, C, and E.
3. Only E.
4. Only D.
136
MAT2611/101/3/2017
Question 15

1 −1 2
Which of the following sets are contained in (i.e. subset of) the column space of ?
1 2 −1
n T T o
A. 1 1 , −1 2
n T T o
B. −1 2 , 2 1
n T T T o
C. 1 0 , 0 1 , 1 1

D. −1 1 1
1. Only D.
2. Only A, B and C.
3. Only A and B.
4. Only A and C.
Question 16
 
2 −4
Which of the following sets are contained in (i.e. subset of) the column space of −1 2  ?
2 −4
n T T o
A. 2 −1 2 , −4 2 −4

B. 2 −4 , −1 2 , 2 −4

C. 2 −4 , −1 2
n T o
D. 2 −1 2

E. 1 2
1. Only E.
2. Only B, C and D.
3. Only B and C.
4. Only A and D.
137
Question 17

1 −1 2
1 2 −1
n T o
A. 1 1 , −1 2

B. 7 −7 −7

C. −1 1 1
1. Only B and C.
2. Only B.
3. Only C.
4. Only A.
Question 18
 
2 −4
Which of the following sets are a basis for the null space of −1 2  ?
2 −4
n T T o
A. 2 −1 2 , −4 2 −4

B. 2 −4 , −1 2

C. 1 2

D. 2 −1
1. Only C and D.
2. Only D.
3. Only B.
4. Only A.
138
MAT2611/101/3/2017
Question 19
Which of the following statements are always true for for all m, n ∈ N and m × n matrix A ?
A. nullity(A) = nullity(AT )
B. rank(AT ) + nullity(A) = m
C. rank(AT ) + nullity(A) = n
1. Only A.
2. Only B.
3. Only C and D.
4. Only C.
Question 20
1. Only A.
2. Only C and D.
3. Only C.
4. Only B.
139
Question 21
1. λ = 0.
4. A is invertible.
Question 22
2. λ = 0.
3. A is invertible.
Question 23
Let A be an n × n matrix with eigenvalue 2 and let I be the n × n identity matrix. Which of the
following are true?
A. -1 is an eigenvalue of A − 3I.
B. rank(A + 3I) = n.
C. 8 is an eigenvalue of A3 .
D. 6 is an eigenvalue of 3A.
1. Only B, C and D.
2. Only B.
3. Only A, C and D.
4. Only C.
140
MAT2611/101/3/2017
Question 24
following are true?
A. 4 is an eigenvalue of A + I.
B. A + 3I is invertible.
1. Only A, C and D.
2. Only B.
3. Only B, C and D.
4. Only C and D.
Question 25

0 0 −1 0 1 2 0 0
A. . B. . C. . D. .
1 0 1 −1 1 2 1 1
1. Only C and D.
2. Only C.
3. Only B.
4. Only A and C.
Question 26

0 1 2 1 0 0 1 1
A. . B. . C. . D. .
0 0 0 2 1 0 1 1
141
1. Only A and C.
2. Only B.
3. Only B and D.
4. Only D.
Question 27
Let A be an n × n matrix and let I be the n × n identity matrix. Then

4. If A is diagonalizable then A + xI is diagonalizable for all x ∈ R.
Question 28

Question 29

1 1 T
1 1
2. h(x1 , x2 ), (y1 , y2 )i = 2x1 y1 + 3x2 y2 in R2 .
3. h(x1 , x2 ), (y1 , y2 )i = (x1 + x2 )(y1 + y2 ) in R2 .
4. h(x1 , x2 ), (y1 , y2 )i = x1 y1 + x2 y2 + 1 in R2 .
142
MAT2611/101/3/2017
Question 30

1. h(x1 , x2 ), (y1 , y2 )i = (x1 + x2 )(y1 + y2 ) in R2 .
2. h(x1 , x2 ), (y1 , y2 )i = x1 y1 − x2 y2 in R2 .

0 1 0 1
3. hA, Bi = tr AB T in M22 .
1 0 1 0
4. hA, Bi = tr (AB) in M22 .
Question 31
h(x1 , x2 , x3 ), (y1 , y2 .y3 )i = 2x1 y1 + 2x2 y2 + 2x3 y3 in R3 ?
√
A. (1, 0, 0) B. (0, 1, 0) C. (1, 0, 0)/ 2 D. (1, 1, 0)/2
1. Only A and B.
2. Only A and C.
3. Only C and D.
4. Only A, B and D.
Question 32
h(x1 , x2 , x3 ), (y1 , y2 .y3 )i = x1 y1 + x2 y2 + 7x3 y3 in R3 ?
√
A. (1, 0, 0) B. (0, 0, 1) C. (1, 1, 1)/ 3 D. (1, 1, 1)/3
1. Only A and B.
2. Only A and C.
3. Only A and D.
4. Only A, B and C.
143
Question 33
Which of the following vectors are orthogonal to each other with respect to the inner product
hA, Bi = tr(AT B) in M22 ?

0 1 0 1 0 1 1 1
A. . B. . C. . D. .
1 0 0 0 −1 0 1 1
1. Only A and B are orthogonal.
2. Only A and C are orthogonal, C and D are orthogonal.
3. Only B and C are orthogonal, B and D are orthogonal.
4. Only A and D are orthogonal.
Question 34
Which of the following vectors are not orthogonal to each other with respect to the inner product

1 0 0 1 0 1 1 1
A. . B. . C. . D. .
0 −1 1 1 −1 0 1 1
1. Only A and B are not orthogonal, B and D are not orthogonal.
2. Only A and C are not orthogonal, B and D are not orthogonal.
3. Only A and C are not orthogonal.
4. Only A and D are not orthogonal.
Question 35
Consider the vector subspace W = span{ 1 + x, 1 + x2 } of P2 with the evaluation inner product at 0,
1 and −1 (sample points). Which of the following vectors in P2 lie in the subspace W ⊥ ?
1. 5x2 + x − 4.
2. x2 + x − 1.
3. −x2 − x + 1.
4. x − 1.
144
MAT2611/101/3/2017
Question 36
Consider the vector subspace W = span{ 1 + x, 1 + x2 } of P2 with the standard inner product. Which
of the following vectors in P2 lie in the subspace W ⊥ ?
1. x2 + x − 1.
2. 5x2 + x − 4.
3. x2 − 2x + 1.
4. x2 − x.
145
Solutions
Question 1
Consider the set

X := { (x, y) : x, y ∈ R }
· : R × X → X, k · a ≡ k · (x, y) := (kx − k + 1, ky),

+ : X × X → X, a + b ≡ (x, y) + (α, β) := (x + α − 1, y + β).
1. (1, 0)
2. (1, 1)
3. (0, 1)
4. (0, 0)
Answer: 1
For all a = (x, y) ∈ X we have
a + (1, 0) = (x, y) + (1, 0) = (x + 1 − 1, y + 0) = (x, y) = a
i.e. 0 = (1, 0) in X. The zero vector is unique (exercise: prove uniqueness of the zero vector using
the existence of the negative of a vector).
Alternative: 0 = 0 · a = 0 · (x, y) = (0x − 0 + 1, 0y) = (1, 0).
Question 2
Consider the set

X := { (x, y) : x, y ∈ R }
· : R × X → X, k · a ≡ k · (x, y) := (kx − k + 1, ky),

+ : X × X → X, a + b ≡ (x, y) + (α, β) := (x + α − 1, y + β).
1. −(1, 0) = (−1, 0)
146
MAT2611/101/3/2017
2. −(1, 0) = (1, 1)
3. −(1, 0) = (1, 0)
4. −(1, 0) = (0, 0)
Answer: 3
Since −(1, 0) = (−1) · (1, 0) we find −(1, 0) = ((−1) · 1 − (−1) + 1, (−1) · 0) = (1, 0).
(This is because (1, 0) is the zero vector in X).
Question 3

A. span { (2, 3) }
B. { (1, x) : x ∈ R }
C. { (0, x) : x ∈ R, x ≥ 0 }
D. { (x, x) : x ∈ R }
1. Only A.
2. Only A and D.
3. Only C.
4. Only C and D.
Answer: 2
The span of a set of element of a vector space is a subspace by definition, so A gives a subspace.
A subspace always includes the zero vector, in this case the zero vector is (0, 0) in R2 , however B
only contains elements of the form (1, x), i.e. the first component of the pair is 1. Thus the set
given in B does not contain the zero vector and is not a subspace of R2 . A subspace must be closed
under scalar multiplication. Note that (0, 1) ∈ { (0, x) : x ∈ R, x ≥ 0 }, but (−1) · (0, 1) = (0, −1) ∈
/
{ (0, x) : x ∈ R, x ≥ 0 } in R2 . So C does not provide a subspace. Now let us consider D.
1. (0, 0) ∈ { (x, x) : x ∈ R }
2. For all a, b ∈ R: (a, a), (b, b) ∈ { (x, x) : x ∈ R } ⇒ (a, a)+(b, b) = (a+b, a+b) ∈ { (x, x) : x ∈ R }
since a + b ∈ R and the first an second component of the pair are identical.
3. For all k, a ∈ R: (a, a) ∈ { (x, x) : x ∈ R } ⇒ k · (a, a) = (ka, ka) ∈ { (x, x) : x ∈ R } since
ka ∈ R and the first an second component of the pair are identical.
Thus D provides a subspace of R2 . (This can also be seen from { (x, x) : x ∈ R } = span{(1, 1)}.)
147
Question 4
A. span { (π, 0) }
B. { (2, x) : x ∈ R }
C. { (x, y) : x, y ∈ N }
D. { (x, −x) : x ∈ R }
1. Only A and D.
2. Only A, B and D.
3. Only C.
4. Only D.
Answer: 1
The span of a set of element of a vector space is a subspace by definition, so A gives a subspace. A
subspace always includes the zero vector, in this case the zero vector is (0, 0) in R2 , however B only
contains elements of the form (2, x), i.e. the first component of the pair is 2. Thus the set given in B
does not contain the zero vector and is not a subspace of R2 . A subspace must be closed under scalar
multiplication. Note that (1, 0) ∈ { (x, y) : x, y ∈ N }, but 0.5 · (1, 0) = (1, 0) ∈
/ { (x, y) : x, y ∈ N } in
R2 . So C does not provide a subspace. Now let us consider D.
1. (0, 0) ∈ { (x, −x) : x ∈ R }
2. For all a, b ∈ R: (a, −a), (b, −b) ∈ { (x, −x) : x ∈ R } ⇒ (a, −a) + (b, −b) = (a + b, −(a + bi)) ∈
{ (x, −x) : x ∈ R } since a + b ∈ R and the first an second component of the pair are negatives
of each other.
3. For all k, a ∈ R: (a, −a) ∈ { (x, −x) : x ∈ R } ⇒ k · (a, −a) = (ka, −ka) ∈ { (x, −x) : x ∈ R }
since ka ∈ R and the first an second component of the pair are negatives of each other.
Thus D provides a subspace of R2 . (This can also be seen from { (x, −x) : x ∈ R } = span{(1, −1)}.)
Question 5
A. span { (2, 3) } in R2
B. { (1, 1), (1, −1) } in R2
148
MAT2611/101/3/2017
C. { (1, 1), (1, −1), (0, 1) } in R2
D. { 1 + x, 1 − x } in P1
1. Only A.
2. Only B.
3. Only B and C.
4. Only B and D.
Answer: 4
By definition, the span of a set (given in A) is linearly dependent. Consider B, solving
c1 (1, 1) + c2 (1, −1) = (c1 + c2 , c1 − c2 ) = (0, 0)
yields c1 = −c2 and c1 = c2 = −c1 , i.e. 2c1 = 0 so that c1 = c2 = 0 which is the only solution. Thus
the set given by B is linearly independent. Consider C, solving
c1 (1, 1) + c2 (1, −1) + c3 (0, 1) = (c1 + c2 , c1 − c2 + c3 ) = (0, 0)
yields c1 = −c2 and c1 = c2 − c3 = −c1 − c3 , i.e. 2c1 = −c3 so that c1 = −c2 = −c3 /2 which yields
infinitely many solutions, for example take c1 = 1, c2 = −1 and c3 = −2. This is a non-zero solution.
Thus the set given by C is linearly dependent. Consider D, solving
c1 (1 + x) + c2 (1 − x) = (c1 + c2 ) · 1 + (c1 − c2 )x = 0 · 1 + 0x
and comparing coefficients in the standard basis {1, x} in P1 yields c1 = −c2 and c1 = c2 = −c1 , i.e.
2c1 = 0 so that c1 = c2 = 0 which is the only solution. Thus the set given by D is linearly independent.
Question 6

A. span { (π, 0) } in R2
B. { (1, 2), (2, 1) } in R2

0 1 0 1 0 0
C. , , in M22
1 0 −1 0 1 0
D. { 1 + 2x, 2 − x } in P1
1. Only B and D.
149
2. Only B.
3. Only C and D.
4. Only D.
Answer: 1
By definition, the span of a set (given in A) is linearly dependent. Consider B, solving
c1 (1, 2) + c2 (2, 1) = (c1 + 2c2 , 2c1 + c2 ) = (0, 0)
yields c1 = c2 , i.e. 3c1 = 0 so that c1 = c2 = 0 which is the only solution. Thus the set given by B is
linearly independent. Consider C, solving

0 1 0 1 0 0 0 c1 + c2 0 0
c1 + c2 + c3 = =
1 0 −1 0 1 0 c1 − c2 + c3 0 0 0
yields c1 = −c2 and c3 = 2c2 which yields infinitely many solutions, for example take c1 = 1, c2 = −1
and c3 = −2. This is a non-zero solution. Thus the set given by C is linearly dependent. Consider D,
solving
c1 (1 + 2x) + c2 (2 − x) = (c1 + 2c2 ) · 1 + (2c1 − c2 )x = 0 · 1 + 0x
and comparing coefficients in the standard basis {1, x} in P1 yields c1 = −2c2 and c2 = 2c1 = −4c2 so
that c1 = c2 = 0 which is the only solution. Thus the set given by D is linearly independent.
Question 7
A. span { (1, 0, 1), (1, 0, −1) } in R3
B. span { (0, 0, 1), (5, 0, 0) } in R3
C. span { (3, 0, 7), (0, 0, 1), (5, 0, 0) } in R3
D. { (1, 0, 1), (1, 0, −1) }
E. span { (1, 1, 1), (1, −1, −1) } in R3
1. Only A and D.
2. Only A and E.
3. Only A, B and C.
4. Only B and C.
150
MAT2611/101/3/2017
Answer: 3
The sets in A, B, C and E have infinitely many elements (due to the span) while the set in D has
only two elements and so cannot be equal to any of the other sets. Now we compare A and B. If each
element of the set in A (respectively B) can be expressed as a linear combination of elements in B
(respectively A) then the two sets are equal:
(1, 0, 1) = a1 (0, 0, 1) + b1 (5, 0, 0) ⇒ a1 = 1, b1 = 1/5

(1, 0, −1) = a2 (0, 0, 1) + b2 (5, 0, 0) ⇒ a2 = −1, b2 = 1/5
(0, 0, 1) = a3 (1, 0, 1) + b3 (1, 0, −1) ⇒ a3 = 1/2, b3 = −1/2
(5, 0, 0) = a4 (1, 0, 1) + b4 (1, 0, −1) ⇒ a4 = 1/2, b4 = −1/2
Since we found a solution in each case, the two sets are equal. Now consider, B and C:
(0, 0, 1) = a1 (3, 0, 7) + b1 (0, 0, 1) + c1 (5, 0, 0) ⇒ b1 = 1 − 7a1 , c1 = −3a1 /5

(5, 0, 0) = a2 (3, 0, 7) + b2 (0, 0, 1) + c2 (5, 0, 0) ⇒ b2 = −7a2 , c2 = 1 − 3/5a2
(3, 0, 7) = a3 (0, 0, 1) + b3 (5, 0, 0) ⇒ a3 = 7, b3 = 5/3
(0, 0, 1) = a4 (0, 0, 1) + b4 (5, 0, 0) ⇒ a4 = 1, b4 = 0
(5, 0, 0) = a5 (0, 0, 1) + b5 (5, 0, 0) ⇒ a5 = 0, b5 = 1
It is sufficient to note that a1 = 0, b1 = 1, c1 = 0, a2 = 0, b2 = 0 and c2 = 1 satisfy the first two

equations (instead of providing all solutions as above). Since we found a solution in each case, the
two sets are equal. Hence, the sets given in A, B and C are all equal. Now consider A and E:
(1, 0, 1) = a(1, 1, 1) + b(1, −1, −1) ⇒ no solution for a, b ∈ R
so that the set in E is not equal to any of the other sets.
Question 8

A. span { (1, 0, 1), (2, 0, 2) } in R3
B. span { (1, 0, 1), (1, 0, −1) } in R3
C. span { (1, 0, 1), (1, 1, 1) } in R3
D. span { (3, 0, 3) } in R3
E. { (1, 0, 1), (2, 0, 2) }

1. Only A and B.
2. Only A and D.
151
3. Only A and E.
4. Only B and C.
Answer: 2
The sets in A, B, C and D have infinitely many elements (due to the span) while the set in E has
only two elements and so cannot be equal to any of the other sets. Now we compare A and B. If each
element of the set in A (respectively B) can be expressed as a linear combination of elements in B
(respectively A) then the two sets are equal:
(1, 0, 1) = a1 (1, 0, 1) + b1 (1, 0, −1) ⇒ a1 = 1, b1 = 0

(2, 0, 2) = a2 (1, 0, 1) + b2 (1, 0, −1) ⇒ a2 = 2, b2 = 0
(1, 0, 1) = a3 (1, 0, 1) + b3 (2, 0, 2) ⇒ a3 = 1, b3 = 0 (amongst others)
(1, 0, −1) = a4 (1, 0, 1) + b4 (2, 0, 2) ⇒ no solution
The two sets are not equal. Now consider, A and C (we show only one equation which provides no
solution):
(1, 1, 1) = a5 (1, 0, 1) + b5 (2, 0, 2) ⇒ no solution
The two sets are not equal. Now consider, B and C (we show only one equation which provides no
solution):
(1, 1, 1) = a6 (1, 0, 1) + b6 (1, 0, −1) ⇒ no solution
Thus A, B and C are all pair wise unequal. Similarly B, C and D are all pair wise unequal. Finally,
consider A and D
(1, 0, 1) = a7 (3, 0, 3) ⇒ a7 = 1/3

(2, 0, 2) = a8 (3, 0, 3) ⇒ a8 = 2/3
(3, 0, 3) = a9 (1, 0, 1) + b9 (2, 0, 2) ⇒ a9 = 3, b9 = 0 (amongst others)
so that A and D are equal.
Question 9
Which of the following sets are a basis for the following vector subspace of P2 :
X = { p(x) ∈ P2 : p(3) = 0 } .
A. { 1, x, x2 }
B. { x − 3, x2 − 9 }
C. { x2 + 2x − 15, x2 − 2x − 3 }
152
MAT2611/101/3/2017
D. { x − 3, x3 − 27 }
1. Only B and C.
2. Only B.
3. Only A.
4. Only B and D.
Answer: 1
Since 1 6= 0, we have 1|x=3 = 1 ∈ / X, x2 |x=3 = 9 6= 0, x2 ∈

/ X (also x|x=3 = 3 6= 0, x ∈ / X). Thus
A does not provide a basis for X. Since X ⊆ P2 , X does not include any cubic polynomials and D
also does not provide a basis for X. Since all the polynomials which appear in B and C are either
quadratic or linear, they are all in P2 . We also have
(x − 3)|x=3 = (x2 − 9)|x=3 = (x2 + 2x − 15)|x=3 = (x2 − 2x − 3)|x=3 = 0
so that { x − 3, x2 − 9 } ⊆ X and { x2 + 2x − 15, x2 − 2x − 3 } ⊆ X. We describe X more explicitly:
X = { p(x) ∈ P2 : p(3) = 0 }
= ax2 + bx + c : a, b, c ∈ R, (ax2 + bx + c)|x=3 = 0

= ax2 + bx + c : a, b, c ∈ R, 9a + 3b + c = 0

= ax2 + bx − 9a − 3b : a, b ∈ R .

Now we need to determine, for B and C, whether each set is linearly dependent and spans X. For B
we have
c1 (x − 3) + c2 (x2 − 9) = 0 + 0 · x + 0 · x2 ⇔ (−3c1 − 9c2 ) + c1 x + c2 x2 = 0 + 0 · x + 0 · x2
⇔ c1 = c2 = 0
so that the set in B is linearly independent (here we used the fact that {1, x, x2 } is a basis for P2 and
compared coefficients). We also have
span x − 3, x2 − 9 = a(x − 3) + b(x2 − 9) : a, b ∈ R

= bx2 + ax − 3a − 9b : a, b ∈ R

= ax2 + bx − 9a − 3b : a, b ∈ R

=X
so that { x − 3, x2 − 9 } spans X. Thus B provides a basis for X. For C we have
c1 (x2 + 2x − 15) + c2 (x2 − 2x − 3) = 0 + 0 · x + 0 · x2
⇔ (−15c1 − 3c2 ) + (2c1 − 2c2 )x + (c1 + c2 )x2 = 0 + 0 · x + 0 · x2
⇔ 5c1 + 3c2 = c1 + c2 = c1 − c2 = 0
⇔ c1 = c2 = 0
153
so that the set in C is linearly independent (here we used the fact that {1, x, x2 } is a basis for P2 and
compared coefficients). We also have
span x2 + 2x − 15, x2 − 2x − 3 = a(x2 + 2x − 15) + b(x2 − 2x − 3) : a, b ∈ R

= (a + b)x2 + 2(a − b)x − 15a − 3b : a, b ∈ R

= a0 x2 + b0 x − 9a0 − 3b0 : a0 , b0 ∈ R

=X
where we set a0 = a + b and b0 = 2(a − b) so that a = a0 /2 + b0 /4 and b = a0 /2 − b0 /4. Note that
{a + b : a, b ∈ R} = R and {2(a − b) : a, b ∈ R} = R. Thus { x2 + 2x − 15, x2 − 2x − 3 } spans X
and C also provides a basis for X.
Question 10

1 0
X = A ∈ M22 : A = .
2 0

2 0 0 −1 0 0 0 0
A. , , ,
0 0 0 0 2 0 0 −1

2 −1
B.
2 −1

2 −1 0 0
C. ,
0 0 2 −1

2 −1 2 −1
D. ,
2 −1 −2 1
1. Only A.
2. Only B.
3. Only C.
4. Only C and D.
Answer: 4
A basis consists of elements of the vector space for which it is a basis. Since [ 20 00 ] ∈
/ X, A cannot
describe a basis for X. Theremaining sets are all subsets of X (check this yourself). Since 42 −1

−2 ∈ X
2 −1
but 24 −1

−2 ∈ / span{ }, B does not provide a basis for X. We notice that
2 −1

a b 1 a + 2b 0
= = ⇔ a = −2b, c = −2d
c d 2 c + 2d 0
154
MAT2611/101/3/2017
so that

−2b b −2 1 0 0 −2 1 0 0
X= : b, d ∈ R = b +d : b, d ∈ R = span , .
−2d d 0 0 −2 1 0 0 −2 1
Since the matrices in C span X and are linearly independent (check this yourself) this set forms a
basis for X. Thus X is two dimensional. Since the set given in D is linearly independent (check this
yourself), and consists of 2 elements of X, it must be a basis for X (since every linearly independent
2-element subset of X is a basis for X). Alternatively,

2 −1 2 −1
span ,
2 −1 −2 1

2 −1 2 −1
= a +b : a, b ∈ R
2 −1 −2 1

2 −1 0 0
= (a + b) + (a − b) : a, b ∈ R
0 0 2 −1

0 2 −1 0 0 0 0 0
= a +b : a ,b ∈ R (set a0 = a + b and b0 = a − b)
0 0 2 −1

−2 1 0 0
= span , = X.
0 0 −2 1
The second last equivalence follow by noting that if given two values α, β ∈ R, a = (α + β)/2 and
b = (α − β)/2 provides a0 = α and b0 = β.
Question 11
A. dim(span { (1, 0, 1), (1, 0, −1) }) = 2 in R3
B. dim(span { (0, 0, 0), (5, 0, 0) }) = 2 in R3
C. dim(span { (3, 0, 7), (0, 0, 1), (5, 0, 0) }) = 2 in R3
D. dim(span { (3, 0, 7), (0, 0, 1), (5, 0, 0) }) = 3 in R3
2. Only A and B.
3. Only A and C.
4. Only A and D.
155
Answer: 3
Since { (1, 0, 1), (1, 0, −1) } consists of two linearly independent vectors (prove this!) A is true. Since
span { (0, 0, 0), (5, 0, 0) } = span { (5, 0, 0) } which has dimension 1, B is false. Now consider C:
c1 (3, 0, 7) + c2 (0, 0, 1) + c3 (5, 0, 0) = (0, 0, 0)
yields 3c1 + 5c3 = 0 and 7c1 + c2 = 0. We apply row reduction to find all solutions:

3 0 5 : 0 3 0 5 : 0
→ (R2 ← 3R2 − 7R1 )
7 1 0 : 0 0 3 −35 : 0
so that c1 = −5c3 /3, c2 = 35c3 /3 and c3 is arbitrary (free). Thus { (3, 0, 7), (0, 0, 1), (5, 0, 0) } is not a
linearly independent set. However { (0, 0, 1), (5, 0, 0) } is linearly independent, so that
span { (3, 0, 7), (0, 0, 1), (5, 0, 0) } = span { (0, 0, 1), (5, 0, 0) }
and C is true. Consequently, D is false.
Question 12
A. dim(span { (1, 0, 1), (1, 0, −1) }) = 2 in R3
B. dim(span { (0, 0, 0), (5, 0, 0) }) = 2 in R3
C. dim(span { (1, 0, 1), (1, 0, −1), (5, 0, 0) }) = 2 in R3
D. dim(span { (1, 0, 1), (1, 0, −1), (5, 0, 0) }) = 3 in R3
2. Only A, B and D.
3. Only A and C.
4. Only A and D.
Answer: 3
Since { (1, 0, 1), (1, 0, −1) } consists of two linearly independent vectors (prove this!) A is true. Since
span { (0, 0, 0), (5, 0, 0) } = span { (5, 0, 0) } which has dimension 1, B is false. Now consider C:
c1 (1, 0, 1) + c2 (1, 0, −1) + c3 (5, 0, 0) = (0, 0, 0)
yields c1 + c2 + 5c3 = 0 and c1 − c2 = 0. Since c1 = c2 = 5 and c3 = −2 provides a non-zero solution.
Thus the dimension is less than 3 (and D is false).
c1 (1, 0, 1) + c2 (1, 0, −1) = (0, 0, 0)
however does yield only the trivial solution, so the dimension is at least 2. Thus the dimension must
be 2.
156
MAT2611/101/3/2017
Question 13

1 −1 2
Which of the following sets are a basis for the row space of ?
1 2 −1

A. 0 −3 3 , 1 2 −1

B. −1 1 1 , 0 0 1

C. 1 −1 2 , 1 2 −1
1. Only A.
2. Only C.
3. Only A and C.
4. Only B.
Answer: 3
In order for each set to be a basis for the row space, it must be linearly independent and each element
must be expressible as a linear combination of the rows. For A we have

c1 0 −3 3 + c2 1 2 −1 = 0 0 0 >
Thus c2 = 0 and −3c1 + 2c2 = 0 so that c1 = 0. Since c1 = c2 = 0 is the only solution the set in A is
linearly independent. Now we solve

1 −1 2 = a1 0 −3 3 + b1 1 2 −1

1 2 −1 = a2 0 −3 3 + b2 1 2 −1
which yields b1 = 1, a1 = 1, b2 = 1 and a2 = 0. Thus A provides a basis for the row space. It is
straightforward to verify that the set in B is linearly independent, and solving

1 −1 2 = a1 −1 1 1 + b1 0 0 1

1 2 −1 = a2 −1 1 1 + b2 0 0 1
yields a1 = −1, b1 = 3, but no solution for a2 . Thus B does not provide a basis. The set given in C
is exactly the rows of the matrix. It suffices to check for linear independence:

c1 1 −1 2 + c2 1 2 −1 = 0 0 0
157
yields c1 + c2 = 0, −c1 + 2c2 = 0 and 2c1 − c2 = 0. We apply row reduction to find all solutions:
   
1 1 : 0 1 1 : 0
−1 2 : 0 → 0 2 : 0 (R2 ← R2 + R1 )
2 −1 : 0 2 −1 : 0
 
1 1 : 0
→ 0 2 : 0 (R3 ← R3 − 2R1 )
0 −3 : 0
 
1 1 : 0
→ 0 2 : 0 (R3 ← 2R3 + 3R1 )
0 0 : 0
 
1 1 : 0
→ 0 1 : 0 (R2 ← R2 /2)
0 0 : 0
 
1 0 : 0
→ 0 1 : 0 (R1 ← R1 − R2 )
0 0 : 0
so that c1 = 0 and c2 = 0 is the only solution. Thus C provides a basis for the row space.
Question 14
 
2 −4
Which of the following sets are a basis for the row space of −1 2  ?
2 −4
n T T o
A. 2 −1 2 , −4 2 −4

B. 2 −4 , −1 2 , 2 −4

C. 2 −4 , −1 2

D. 1 2

E. 2 −1
1. Only A.
2. Only B, C, and E.
3. Only E.
4. Only D.
158
MAT2611/101/3/2017
Answer: 5
The row space consists of 2 coordinate row vectors, so A and D make no sense. B is not a basis since
the third vector is a (trivial) linear combination of the first two. C is not a basis since the first vector
is a scalar multiple (i.e. −2) the second vector. The set E lists a vector which cannot be expressed as
a linear combination of the rows:

1 2 = a 2 −4 + b −1 2 + c 2 −4 = 2a − b + 2c −4a + 2b − 4c
The equations
2a − b + 2c = 1
−4a + 2b − 4c = 2
Adding twice the first equation to the second yield 0 = 3, obviously a contradiction.
Question 15

1 −1 2
Which of the following sets are contained in (i.e. subset of) the column space of ?
1 2 −1
n T T o
A. 1 1 , −1 2
n T T o
B. −1 2 , 2 1
n T T T o
C. 1 0 , 0 1 , 1 1

D. −1 1 1
1. Only D.
2. Only A, B and C.
3. Only A and B.
4. Only A and C.
Answer: 2
There are 2 rows, so D is not relevant. The set in A consists of columns of the matrix and is by
T
definition in the column space of the matrix. Similarly the first element −1 2 in the set given in
B is in the column space of the matrix, but we must still check the second element:

2 1 −1 2
=a +b +c
1 1 2 −1
159
yields a − b + 2c = 2 and a + 2b − c = 1 and using row reduction we find

1 −1 2 : 2 1 −1 2 : 2
→ (R2 ← R2 − R1 )
1 2 −1 : 1 0 3 −3 : −1

1 −1 2 : 2
→ (R2 ← R2 /3)
0 1 −1 : −1/3

1 0 1 : 5/3
→ (R1 ← R1 + R2 )
0 1 −1 : −1/3
T
so that we have a solution (a = 5/3 − c, b = c − 1/3 and c is free). Thus 2 1 is in the column
space and the set given in B is a subset of the column space. Similarly, for C we need only check the
first two elements

1 1 −1 2
= a1 + b1 + c1
0 1 2 −1

0 1 −1 2
= a2 + b2 + c2
1 1 2 −1
with solution a1 = 2/3 − c1 , b1 = c1 − 1/3, a2 = 1/3 − c2 and b2 = 1/3 + c2 . It follows that the set
given in C is a subset of the column space.
A much simpler solution is found by noting that the column space is the vector space of column vectors
R2 (since two of the columns are linearly independent). Then the sets in A, B and C are obviously
subsets of the column space.
Question 16
 
2 −4
Which of the following sets are contained in (i.e. subset of) the column space of −1 2  ?
2 −4
n T T o
A. 2 −1 2 , −4 2 −4

B. 2 −4 , −1 2 , 2 −4

C. 2 −4 , −1 2
n T o
D. 2 −1 2

E. 1 2
1. Only E.
2. Only B, C and D.
3. Only B and C.
160
MAT2611/101/3/2017
4. Only A and D.
Answer: 4
There are 3 rows, so B, C and E is not relevant. The set in A consists of columns of the matrix and
is by definition in the column space of the matrix. Similarly the only element in the set given in D is
in the column space of the matrix.
Question 17

1 −1 2
1 2 −1
n T o
A. 1 1 , −1 2

B. 7 −7 −7

C. −1 1 1
1. Only B and C.
2. Only B.
3. Only C.
4. Only A.
Answer: 1
The null space is given by

       
 a a   a 
 b  : a, b, c ∈ R, 1 −1 2  b  = 0 =  b  : a, b, c ∈ R,
a − b + 2c
=
0
1 2 −1 0 a + 2b − c 0 
c c c
  
  
 a 
=  b : c ∈ R, a = −c, b = c,

c
 
  
 −c 
=  c  : c∈R
c
 
 
 −1 
= span  1  .
1
 
161
T T n T o
Since −1 1 1 , 7 −7 −7 ∈ span −1 1 1 are both non-zero vectors in a one dimen-
sional vector space (the null space) both B and C provide a basis for the null space.
Question 18
 
2 −4
Which of the following sets are a basis for the null space of −1 2  ?
2 −4
n T T o
A. 2 −1 2 , −4 2 −4

B. 2 −4 , −1 2

C. 1 2

D. 2 −1

1. Only C and D.
2. Only D.
3. Only B.
4. Only A.
Answer: 5
The null space consists of matrix multiplication compatible vectors, i.e. since the matrix is 3 × 2 -
which means that the null space consists of 2 × 1 column vectors. None of these sets consist of 2 × 1
matrices.
Question 19

1. Only A.
162
MAT2611/101/3/2017
2. Only B.
3. Only C and D.
4. Only C.
Answer: 4
Statement A is false (for some matrices). For example, consider A = [ 10 00 00 ] so that nullity(A) =
2 and nullity(AT ) = 1. Statement B is false (for some matrices) since for the previous exam-
ple rank(AT ) + nullity(A) = 1 + 2 = 3 6= 2 which is the number of rows (m). We know that
rank(A) + nullity(A) = n and that rank(A) = rank(AT ). Thus rank(AT ) + nullity(A) = n is true for
all m × n matrices A. Since the number of rows and columns in A may be different, statement D
makes no sense in general, and is false for some matrices.
Challenge: for each of A, B and D, characterize all the matrices for which the statement is true.
Question 20
1. Only A.
2. Only C and D.
3. Only C.
4. Only B.
Answer: 3
Statement A is false (for some matrices). For example, consider A = [ 10 00 00 ] so that nullity(A) =
2 and nullity(AT ) = 1. Statement B is false (for some matrices) since for the previous exam-
ple rank(AT ) + nullity(A) = 1 + 2 = 3 6= 2 which is the number of rows (m). We know that
rank(A) + nullity(A) = n and that rank(A) = rank(AT ). Thus rank(AT ) + nullity(A) = n is true for
all m × n matrices A. Since the number of rows and columns in A may be different, statement D
makes no sense in general, and is false for some matrices.
Challenge: for each of A, B and D, characterize all the matrices for which the statement is true.
163
Question 21
1. λ = 0.
4. A is invertible.
Answer: 3
First note that x = 0 always satisfies the equation. If λ = 0, and the solution x = 0 is unique, then
A must be invertible – but this was not given. If λ = 0 is an eigenvalue of A, then there exists a
non-zero vector x 6= 0 such that Ax = 0 which contradicts the uniqueness of the solution x = 0. If
λ is and eigenvalue of A then Ax = λx has a non-zero solution for x (contradicting the uniqueness
of the solution x = 0), thus λ is not an eigenvalue of A. Conversely, if λ is not an eigenvalue of
A, then no non-zero vector x exists which satisfies Ax = λx (ensuring the uniqueness of the so-
lution x = 0). Thus 3 is valid. Consider A = I and λ = 1, then A is invertible but any non-zero
vector x satisfies Ax = λx. Thus invertibility is not necessary for the uniqueness of the solution x = 0.
This question emphasizes the fact that no eigenvalue of A has an eigenspace consisting of only the
zero vector. If a calculation to find an eigenspace for the “eigenvalue” λ yields only the zero vector,
either the calculation is incorrect or λ is not an eigenvalue.
Question 22
2. λ = 0.
3. A is invertible.
Answer: 1
If λ is and eigenvalue of A then Ax = λx has a non-zero solution for x (contradicting the uniqueness
of the solution x = 0), thus λ is not an eigenvalue of A. Conversely, if λ is not an eigenvalue of A,
164
MAT2611/101/3/2017
then no non-zero vector x exists which satisfies Ax = λx (ensuring the uniqueness of the solution
x = 0). Thus 1 is valid. If λ = 0, and the solution x = 0 is unique, then A must be invertible – but
this was not given. Consider A = I and λ = 1, then A is invertible but any non-zero vector x satisfies
Ax = λx. Thus invertibility is not necessary for the uniqueness of the solution x = 0. If λ = 0 is an
eigenvalue of A, then there exists a non-zero vector x 6= 0 such that Ax = 0 which contradicts the
uniqueness of the solution x = 0.
This question emphasizes the fact that no eigenvalue of A has an eigenspace consisting of only the
zero vector. If a calculation to find an eigenspace for the “eigenvalue” λ yields only the zero vector,
either the calculation is incorrect or λ is not an eigenvalue.
Question 23
following are true?
A. -1 is an eigenvalue of A − 3I.
B. rank(A + 3I) = n.
1. Only B, C and D.
2. Only B.
3. Only A, C and D.
4. Only C.
Answer: 3
Since 2 is an eigenvalue of A, there exists a non-zero vector x ∈ Rn such that Ax = 2x. Now
(A − 3I)x = Ax − 3x = 2x − 3x = −x. Since x is non-zero, it is clear that -1 is an eigenvalue of
A − 3I with corresponding eigenvector x. Thus A is true. Let
 
2 0 0 0 ···
0 −3 0 0 · · ·
A = 0 0 −3 0 · · · .
 
 
.. .. . . . . . .
. . . . .
Then 2 is an eigenvalue of A, but rank(A + 3I) = 1 and B is false in general. We have A3 x =

A(A(Ax)) = A(A(2x)) = 2(A(Ax)) = 4(Ax) = 8x and since x 6= 0, C is true. Also, 3Ax = 3(Ax) =
6x so that D is true.
165
Question 24
following are true?
A. 4 is an eigenvalue of A + I.
B. A + 3I is invertible.
1. Only A, C and D.
2. Only B.
3. Only B, C and D.
4. Only C and D.
Answer: 1
Since 3 is an eigenvalue of A, there exists a non-zero vector x ∈ Rn such that Ax = 3x. Now
(A + I)x = Ax + x = 3x + x = 4x. Since x is non-zero, it is clear that 4 is an eigenvalue of A + I
with corresponding eigenvector x. Thus A is true. If A = −3I then A + 3I is obviously not invertible
and B is false in general. We have A2 x = A(Ax) = A(3x) = 3(Ax) = 9x and since x 6= 0, C is true.
Also, 2Ax = 2(Ax) = 6x so that D is true.
Question 25

0 0 −1 0 1 2 0 0
A. . B. . C. . D. .
1 0 1 −1 1 2 1 1
1. Only C and D.
2. Only C.
3. Only B.
4. Only A and C.
166
MAT2611/101/3/2017
Answer: 1
The matrix D is lower triangular, the eigenvalues lie on the diagonal i.e. 0 and 1. Since all of the
eigenvalues are distinct, D is diagonal. We consider the eigenvalues and eigenspaces for the matrices
in A and B. These matrices are all upper or lower triangular, which means the eigenvalues are given
by the diagonal entries of the matrices. Since the matrices A and B have only one eigenvalue with
algebraic multiplicity 2, we need only check whether the geometric multiplicity is also 2 for each case.
Matrix Eigenvalue Algebraic Eigenspace Geometric Diagonalizable
multiplicity multiplicity

0 0 0
0 2 span 1 No
1 0 1

−1 0 0
−1 2 span 1 No
1 −1 1
The last matrix (C) has characteristic equation
λ2 − 3λ = λ(λ − 3) = 0
so that C has two distinct eigenvalues and is diagonalizable.
Question 26

0 1 2 1 0 0 1 1
A. . B. . C. . D. .
0 0 0 2 1 0 1 1
1. Only A and C.
2. Only B.
3. Only B and D.
4. Only D.
Answer: 4
The matrix in D is symmetric and therefore diagonalizable. We consider the eigenvalues and eigenspaces
for the matrices in A, B and C. These matrices are all upper or lower triangular, which means the
eigenvalues are given by the diagonal entries of the matrices. Since each matrix has only one eigen-
value with algebraic multiplicity 2, we need only check whether the geometric multiplicity is also 2
for each case.
167
Matrix Eigenvalue Algebraic Eigenspace Geometric Diagonalizable
multiplicity multiplicity

0 1 1
0 2 span 1 No
0 0 0

2 1 1
2 2 span 1 No
0 2 0

0 0 0
0 2 span 1 No
1 0 1
Question 27

Answer: 4
The n × n zero matrix is trivially diagonalizable (it is already diagonal) but is not invertible. Thus 1
is false. The n × n zero matrix is trivially diagonalizable
−1 0 (it is already diagonal) with all eigenvalues
equal to zero. Thus 2 is false. The matrix 1 −1 (from above) is invertible but not diagonalizable,
hence 3 is false (can you think of a counter example for 3 × 3 matrices? n × n matrices?). If A is
diagonalizable, then there exists an invertible n × n matrix P such that P −1 AP is diagonal. Now,
P 1 (A + xI)P = P −1 AP + xP P −1 = P −1 AP + xI is the sum of two diagonal matrices which is also
diagonal. Thus A + xI is diagonalizable (diagonalized by P ). Thus 4 is true.
Question 28

168
MAT2611/101/3/2017
Answer: 1
If A is diagonalizable, then there exists an invertible n × n matrix P such that P −1 AP is diagonal.

Now, P 1 (A + xI)P = P −1 AP + xP P −1 = P −1 AP + xI is the sum of two diagonal matrices which
is also diagonal. Thus A + xI is diagonalizable (diagonalized by P ). Thus 1 is true. The n × n zero
matrix is trivially diagonalizable (it is already diagonal) but is not invertible. Thus 2 is false. The
matrix [ 20 12 ] (from above) is invertible but not diagonalizable, hence 3 is false (can you think of a
counter example for 3 × 3 matrices? n × n matrices?). The n × n zero matrix is trivially and has the
eigenvalue 0. Thus 5 is false.
Question 29

1 1 T
1 1
2. h(x1 , x2 ), (y1 , y2 )i = 2x1 y1 + 3x2 y2 in R2 .
3. h(x1 , x2 ), (y1 , y2 )i = (x1 + x2 )(y1 + y2 ) in R2 .
4. h(x1 , x2 ), (y1 , y2 )i = x1 y1 + x2 y2 + 1 in R2 .
Answer: 2
1 1
1. Consider A = [ −1 −1 ], then hA, Ai = 0 which violates positivity.
2. This is an example of a weighted Euclidean inner product (see the text book).
3. Consider h(1, −1), (1, −1)i = 0 which violates positivity.
4. Here h(0, 0), (0, 0)i = 1 6= 0.
Question 30
1. h(x1 , x2 ), (y1 , y2 )i = (x1 + x2 )(y1 + y2 ) in R2 .
2. h(x1 , x2 ), (y1 , y2 )i = x1 y1 − x2 y2 in R2 .

0 1 T 0 1
1 0 1 0
169
4. hA, Bi = tr (AB) in M22 .
Answer: 3
Consider 3 first. Since

T !
a1 a2 b1 b2 0 1 a1 a2 b1 b2 0 1
, = tr
a3 a4 b3 b4 1 0 a3 a4 b3 b4 1 0

a b + a4 b 4 a3 b1 + a4 b2
= tr 3 3
a1 b 3 + a2 b 4 a1 b 1 + a2 b 2
= a1 b 1 + a2 b 2 + a3 b 3 + a4 b 4
which is the inner product on M22 given as an example in the text book. (This follows trivially from
tr(ABC) = tr(CAB), and the consequence tr(AB) = tr(BA) for product compatible square matrices
A, B and C.) It is left as an exercise to show that each of the axioms hold (note that this inner
product is essentially identical to the standard inner product on R4 . For the remaining cases we
provide examples of axioms which do not hold (other examples also exist).
1. h(1, −1), (1, −1)i = 0 which violates positivity (since (1, −1) 6= (0, 0)).
2. h(1, 1), (1, 1)i = 0 which violates positivity (since (1, 1) 6= (0, 0)).

0 1 0 1
4. , = 0 which violates positivity.
0 0 0 0
Question 31
h(x1 , x2 , x3 ), (y1 , y2 .y3 )i = 2x1 y1 + 2x2 y2 + 2x3 y3 in R3 ?
√
A. (1, 0, 0) B. (0, 1, 0) C. (1, 0, 0)/ 2 D. (1, 1, 0)/2
1. Only A and B.
2. Only A and C.
3. Only C and D.
4. Only A, B and D.
Answer: 3
170
MAT2611/101/3/2017
√ √
A. k(1, 0, 0)k = 2 · 12 + 2 · 02 + 2 · 02 =
2
√ √
B. k(0, 1, 0)k = 2 · 02 + 2 · 12 + 2 · 02 = 2
√
r
2
C. k(1, 0, 0)/ 2k = 2 √12 + 2 · 02 + 2 · 02 = 1
q
1 2 1 2

D. k(1, 1, 0)/2k = 2 2
+2 2
+ 2 · 02 = 1
Question 32
h(x1 , x2 , x3 ), (y1 , y2 .y3 )i = x1 y1 + x2 y2 + 7x3 y3 in R3 ?
√
A. (1, 0, 0) B. (0, 0, 1) C. (1, 1, 1)/ 3 D. (1, 1, 1)/3
1. Only A and B.
2. Only A and C.
3. Only A and D.
4. Only A, B and C.
Answer: 3
√
A. k(1, 0, 0)k = 12 + 02 + 7 · 02 = 1
√ √
B. k(0, 0, 1)k = 02 + 02 + 7 · 12 = 7
√
r
2 2 2
C. k(1, 1, 1)/ 3k = √1 + √1 + 7 √1 = √3
3 3 3 3
q
1 2 1 2 1 2

D. k(1, 1, 1)/3k = 3
+ 3
+7 3
=1
Question 33
Which of the following vectors are orthogonal to each other with respect to the inner product

0 1 0 1 0 1 1 1
A. . B. . C. . D. .
1 0 0 0 −1 0 1 1
171
1. Only A and B are orthogonal.
2. Only A and C are orthogonal, C and D are orthogonal.
3. Only B and C are orthogonal, B and D are orthogonal.
4. Only A and D are orthogonal.
Answer: 2
since

0 1 0 1 0 1 0 1 0 1 1 1
, = 1, , = 0, , = 2,
1 0 0 0 1 0 −1 0 1 0 1 1

0 1 0 1 0 1 1 1
, = −1, , = 1,
0 0 −1 0 0 0 1 1

0 1 1 1
, = 0.
−1 0 1 1
Question 34
Which of the following vectors are not orthogonal to each other with respect to the inner product

1 0 0 1 0 1 1 1
A. . B. . C. . D. .
0 −1 1 1 −1 0 1 1
1. Only A and B are not orthogonal, B and D are not orthogonal.
2. Only A and C are not orthogonal, B and D are not orthogonal.
3. Only A and C are not orthogonal.
4. Only A and D are not orthogonal.
Answer: 1
since

1 0 0 1 1 0 0 1 1 0 1 1
, = −1, , = 0, , = 0,
0 −1 1 1 0 −1 −1 0 0 −1 1 1

0 1 0 1 0 1 1 1
, = 0, , = 3,
1 1 −1 0 1 1 1 1

0 1 1 1
, = 0.
−1 0 1 1
172
MAT2611/101/3/2017
Question 35
Consider the vector subspace W = span{ 1 + x, 1 + x2 } of P2 with the evaluation inner product at 0,
1 and −1 (sample points). Which of the following vectors in P2 lie in the subspace W ⊥ ?
1. 5x2 + x − 4.
2. x2 + x − 1.
3. −x2 − x + 1.
4. x − 1.
Answer: 1
If p(x) ∈ W ⊥ then for all a, b ∈ R
hp(x), a(1 + x) + b(1 + x2 )i = ahp(x), 1 + xi + bhp(x), 1 + x2 i = 0.
By choosing a = 1 and b = 0 we must have hp(x), 1 + xi = 0. By choosing a = 0 and b = 1 we must

have hp(x), 1 + x2 i = 0. Imposing these two conditions yields hp(x), a(1 + x) + b(1 + x2 )i = 0. Thus
it is sufficient to determine whether hp(x), 1 + xi = 0 and hp(x), 1 + x2 i = 0. The evaluation inner
products (using the sample points 0, 1 and −1) are given by
hp(x), 1 + xi = p(0) · [1 + x]x→0 + p(1) · [1 + x]x→1 + p(−1) · [1 + x]x→−1

= p(0) · 1 + p(1) · 2 + p(−1) · 0,
hp(x), 1 + x i = p(0) · [1 + x2 ]x→0 + p(1) · [1 + x2 ]x→1 + p(−1) · [1 + x2 ]x→−1
2
= p(0) · 1 + p(1) · 2 + p(−1) · 2.
Thus we find
1. h5x2 + x − 4, 1 + xi = (−4) · 1 + 2 · 2 + 0 · 0 = 0,
h5x2 + x − 4, 1 + x2 i = (−4) · 1 + 2 · 2 + 0 · 2 = 0.
2. hx2 + x − 1, 1 + xi = (−1) · 1 + 1 · 2 + (−1) · 0 6= 0.
3. h−x2 − x − 1, 1 + xi = (−1) · 1 + (−3) · 2 + (−1) · 0 6= 0.
4. hx − 1, 1 + xi = (−1) · 1 + 0 · 2 + (−2) · 0 6= 0.
Question 36
Consider the vector subspace W = span{ 1 + x, 1 + x2 } of P2 with the standard inner product. Which
of the following vectors in P2 lie in the subspace W ⊥ ?
1. x2 + x − 1.
173
2. 5x2 + x − 4.
3. x2 − 2x + 1.
4. x2 − x.
Answer: 1
If p(x) ∈ W ⊥ then for all a, b ∈ R
hp(x), a(1 + x) + b(1 + x2 )i = ahp(x), 1 + xi + bhp(x), 1 + x2 i = 0.
By choosing a = 1 and b = 0 we must have hp(x), 1 + xi = 0. By choosing a = 0 and b = 1 we must

have hp(x), 1 + x2 i = 0. Imposing these two conditions yields hp(x), a(1 + x) + b(1 + x2 )i = 0. Thus
it is sufficient to determine whether hp(x), 1 + xi = 0 and hp(x), 1 + x2 i = 0. The standard inner
products are given by
1. h−1 + x + x2 , 1 + x + 0x2 i = −1 · 1 + 1 · 1 + 1 · 0 = 0,
h−1 + x + x2 , 1 + 0x + x2 i = −1 · 1 + 1 · 0 + 1 · 1 = 0.
2. h−4 + x + 5x2 , 1 + x + 0x2 i = −4 · 1 + 1 · 1 + 5 · 0 = −3 6= 0.
3. h1 − 2x + x2 , 1 + x + 0x2 i = 1 · 1 + (−2) · 1 + 1 · 0 = −1 6= 0.
4. h0 − x + x2 , 1 + x + 0x2 i = 0 · 1 + (−1) · 1 + 1 · 0 = −1 6= 0.
174
MAT2611/101/3/2017
F.3 2016 Semester 1: Exam

Question paper
This question is a multiple choice question and should be answered in the answer book. Any
rough work should be clearly marked and appear on the last pages of the answer book. Write only
the number for your answer.
(1.1) Consider the set (2)

X := { ♠ }
and the operations (for all k ∈ R and a, b ∈ X)
· : R × X → X, k · a := ♠,
+ : X × X → X, a + b := ♠.
The set X with these definitions of · and + forms a vector space. Which of the following
statements are true in X ?
A. for all x ∈ X: −x = ♠
B. for all x ∈ X: −x = x
C. 0 = 0
D. 0 = (0, 0)
Choose from the following:
1. A
2. B
3. A and B
4. C or D
175
(1.2) Which of the following are subspaces of M22 with the usual operations ? (2)

0 0 0 0
A. span ,
0 1 0 −1

0 a
B. : a≥0
−a 0

a −1
C. : a∈R
0 a
1. Only A.
2. Only A and B.
3. Only B and C.
(1.3) Which of the following sets are linearly independent? (2)

1 1 2 1 0 1
A. , , in M22
0 1 0 1 0 1
B. { (1, 0, 1), (0, 1, 0), (1, 1, −1) } in R3
C. { 1 − x, 1 − x2 , 1 − x + x2 } in P2
1. Only A and C.
2. Only B and C.
3. Only B.
4. Only C.
176
MAT2611/101/3/2017
(1.4) Which of the following sets are a basis for the following vector subspace of M22 : (2)

a b
X= : a, b, c ∈ R .
0 c

1 0 0 1 0 0
A. , ,
0 0 0 0 0 1

1 1 −1 1 1 1
B. , ,
0 1 0 0 0 −1
1. Both A and B.
2. Only A.
3. Only B.
(1.5) Which of the following statements are true: (2)

1 1 2 1 0 1
A. dim span , , = 2 in M22
0 1 0 1 0 1
B. dim (span { (1, 0, 1), (0, 1, 0), (1, 1, −1) }) = 3 in R3
C. dim (span { 1 − x, 1 − x2 , 1 − x + x2 }) = 2 in P2
1. Only A.
2. Only B.
3. Only C.
4. Only A and B.
177
 
1 1 −1
0 1 −1
(1.6) Which of the following sets are a basis for the row space of 
1 0
? (2)
0
1 −2 2

A. 1 1 −1 , 0 1 −1 , 1 0 0

B. 1 1 −1 , 0 1 −1

C. 1 1 −1 , 0 1 −1 , 1 0 0 , 1 −2 2
1. Only A.
2. Only B.
3. Both A and B.
4. Only C.
 
1 1 −1
0 1 −1
(1.7) Which of the following sets are a basis for the null space of 
1 0
? (2)
0
1 −2 2
   
 1 0 
A.  0 , 1
 
0 −1
 
 
 0 
B. 1
1
 
 
 0 
C. 2
2
 
1. Only A.
2. Only B.
3. Both B and C.
178
MAT2611/101/3/2017
 
1 1 −1
0 1 −1
(1.8) Which one of the following statements is true for the matrix A = 
1 0
? (2)
0
1 −2 2
1. rank(A) = 3, nullity(A) = 0.
Consider the vector space M22 .
(2.1) Show that (12)

2 0 T
hA, Bi = tr AB
0 1
is an inner product on M22 .

0 0
(2.2) Prove that if A, B ∈ M22 , where A, B 6= , are orthogonal to each other with re- (6)
0 0
spect to the inner product defined in 2.1 above, then { A, B } is a linearly independent
set.
(2.3) Apply the Gram-Schmidt process to the following subset of M22 : (12)

1 1 1 1 0 1
, ,
0 2 0 0 0 1
to find an orthogonal basis with respect to the inner product defined in 2.1 above for
the span of this subset.
(2.4) Let V be a vector space with zero vector 0 and let h·, ·i denote an inner product on V . (4)
Prove that h0, vi = 0 for all v ∈ V .

1 1 0
A = 0 1 0 .
1 1 0
(3.1) Determine the nullity of A. (2)
179
(3.2) Show that the characteristic equation for the eigenvalues λ of A is given by (3)
λ(λ − 1)2 = 0.
(3.3) Find bases for the eigenspaces of A. (14)
(3.4) For each eigenvalue, determine the algebraic and geometric multiplicity. Is A diagonal- (5)
izable?
(3.5) Prove or disprove: (2)
If B is a 2 × 2 matrix, then B is diagonalizable if and only if B 2 is diagonalizable.
(3.6) Let B be an n × n matrix. Prove that B + B T is diagonalizable. (2)

3 x y
Let T : R → M22 be defined by T (x, y, z) = .
z x
(4.1) Show that T is a linear transformation. (4)
(4.2) Find the matrix representation [T ]B2 ,B1 of T relative to the basis (8)
B1 = { (1, 0, 1), (0, 1, 0), (1, 0, −1) }
in R3 and the basis

1 0 1 0 0 1 0 1
B2 = , , ,
0 1 0 −1 1 0 −1 0
in M22 , ordered from left to right.
(4.3) Determine the range R(T ) of T . Is T onto? In other words, is it true that R(T ) = M22 ? (4)
(4.4) Determine ker(T ) and the nullity of T . (4)
(4.5) Is T one-to-one? Motivate your answer. (2)
180
MAT2611/101/3/2017
Solution
This question is a multiple choice question and should be answered in the answer book. Any
rough work should be clearly marked and appear on the last pages of the answer book. Write only
the number for your answer.

X := { ♠ }
and the operations (for all k ∈ R and a, b ∈ X)
· : R × X → X, k · a := ♠,
+ : X × X → X, a + b := ♠.
The set X with these definitions of · and + forms a vector space. Which of the following
statements are true in X ?
A. for all x ∈ X: −x = ♠
B. for all x ∈ X: −x = x
C. 0 = 0
D. 0 = (0, 0)
Choose from the following:
1. A
2. B
3. A and B
4. C or D

Answer: 3
181
(1.2) Which of the following are subspaces of M22 with the usual operations ? (2)

0 0 0 0
A. span ,
0 1 0 −1

0 a
B. : a≥0
−a 0

a −1
C. : a∈R
0 a
1. Only A.
2. Only A and B.
3. Only B and C.

Answer: 1

1 1 2 1 0 1
A. , , in M22
0 1 0 1 0 1
B. { (1, 0, 1), (0, 1, 0), (1, 1, −1) } in R3
C. { 1 − x, 1 − x2 , 1 − x + x2 } in P2
1. Only A and C.
2. Only B and C.
3. Only B.
4. Only C.

Answer: 2
182
MAT2611/101/3/2017
(1.4) Which of the following sets are a basis for the following vector subspace of M22 : (2)

a b
X= : a, b, c ∈ R .
0 c

1 0 0 1 0 0
A. , ,
0 0 0 0 0 1

1 1 −1 1 1 1
B. , ,
0 1 0 0 0 −1
1. Both A and B.
2. Only A.
3. Only B.

Answer: 1

1 1 2 1 0 1
A. dim span , , = 2 in M22
0 1 0 1 0 1
B. dim (span { (1, 0, 1), (0, 1, 0), (1, 1, −1) }) = 3 in R3
C. dim (span { 1 − x, 1 − x2 , 1 − x + x2 }) = 2 in P2
1. Only A.
2. Only B.
3. Only C.
4. Only A and B.

Answer: 4
183
 
1 1 −1
0 1 −1
(1.6) Which of the following sets are a basis for the row space of 
1 0
? (2)
0
1 −2 2

A. 1 1 −1 , 0 1 −1 , 1 0 0

B. 1 1 −1 , 0 1 −1

C. 1 1 −1 , 0 1 −1 , 1 0 0 , 1 −2 2
1. Only A.
2. Only B.
3. Both A and B.
4. Only C.

Answer: 2
 
1 1 −1
0 1 −1
(1.7) Which of the following sets are a basis for the null space of 
1 0
? (2)
0
1 −2 2
   
 1 0 
A. 0 ,  1 
0 −1
 
 
 0 
B. 1
1
 
 
 0 
C. 2
2
 

1. Only A.
2. Only B.
3. Both B and C.

Answer: 3
184
MAT2611/101/3/2017
 
1 1 −1
0 1 −1
(1.8) Which one of the following statements is true for the matrix A = 
1 0
? (2)
0
1 −2 2
Answer: 4

2 0 T
hA, Bi = tr AB
0 1
is an inner product on M22 .
• For all A, B ∈ M22

2 0 T 2 0 2 0
hA, Bi = tr AB T
= tr BA = tr BA T
= hB, AiX2
0 1 0 1 0 1
where we used Theorem TI and Theorem CT.
• For all k ∈ R and A ∈ M22 we have

2 0 2 0
hkA, Bi = tr (kA)B T
= k tr AB T
= khA, BiX2
0 1 0 1
since tr(kA) = k tr(A).
• For all A, B, C ∈ M22

2 0 T 2 0 T T
hA, B + Ci = tr A(B + C) = tr A(B + C )
0 1 0 1

2 0 T 2 0 T
= tr AB + AC
0 1 0 1

2 0 2 0
= tr AB T
+ tr AC T
= hA, Bi + hA, CiX4
0 1 0 1
since tr(A + B) = tr(A) + tr(B).
185

a b
• Let A = . Then
c d

2 0 2 0 a2 + b2 ac + bd
hA, Ai = tr AA T
= tr = 2a2 +2b2 +c2 +d2 ≥ 0X2
0 1 0 1 ac + bd c2 + d2
and hA, Ai = 0 if and only if a = b = c = d = 0 (since a2 , b2 , c2 , d2 ≥ 0). X2

Alternative:
Note that

a1 a2 b1 b2 2 0 a1 a2 b 1 b 3
, = tr = 2a1 b1 + 2a2 b2 + a3 b3 + a4 b4 .
a3 a4 b3 b4 0 1 a3 a4 b 2 b 4

a1 a2 b1 b2
• For all , ∈ M22
a3 a4 b3 b4

a1 a2 b1 b2
, = 2a1 b1 + 2a2 b2 + a3 b3 + a4 b4
a3 a4 b3 b4

b1 b2 a1 a2
= 2b1 a1 + 2b2 a2 + b3 a3 + b4 a4 = , .
b3 b4 a3 a4

a1 a2
• For all k ∈ R and ∈ M22
a3 a4

a1 a2 b1 b2 ka1 ka2 b1 b2
k , = ,
a3 a4 b3 b4 ka3 ka4 b3 b4
= 2(ka1 )b1 + 2(ka2 )b2 + (ka3 )b3 + (ka4 )b4

a1 a2 b1 b2
= k(2a1 b1 + 2a2 b2 + a3 b3 + a4 b4 ) = k , .
a3 a4 b3 b4

a1 a2 b1 b2 c1 c2
• For all , , ∈ M22
a3 a4 b3 b4 c3 c4

a1 a2 b1 b2 c1 c2 a1 a2 b1 + c 1 b2 + c 2
, + = ,
a3 a4 b3 b4 c3 c4 a3 a4 b3 + c 3 b4 + c 4
= 2a1 (b1 + c1 ) + 2a2 (b2 + c2 ) + a3 (b3 + c3 ) + a4 (b4 + c4 )
= (2a1 b1 + 2a2 b2 + a3 b3 + a4 b4 )
+ (2a1 c1 + 2a2 c2 + a3 c3 + a4 c4 )

a1 a2 b1 b2 a1 a2 c1 c2
= , + , .
a3 a4 b3 b4 a3 a4 c3 c4

a1 a2
• Let A = . Then
a3 a4
hA, Ai = 2a21 + 2a22 + a23 + a24 ≥ 0

and hA, Ai = 0 if and only if a1 = a2 = a3 = a4 = 0 (since a21 , a22 , a23 , a24 ≥ 0).
186
MAT2611/101/3/2017

0 0
(2.2) Prove that if A, B ∈ M22 , where A, B 6= , are orthogonal to each other with re- (6)
0 0
spect to the inner product defined in 2.1 above, then { A, B } is a linearly independent
set.
Suppose c1 A + c2 B = 0 where c1 , c2 ∈ R. Since A and B are orthogonal to each other
we have hA, Bi = hB, Ai = 0.X
c1 A + c2 B = 0 ⇒ hA, c1 A + c2 Bi = hA, 0iX

⇒ c1 hA, Ai + c2 hA, Bi = 0X
⇒ c1 hA, Ai = 0X
⇒ c1 = 0X
since hA, Ai =
6 0. Similarly
c1 A + c2 B = 0 ⇒ hB, c1 A + c2 Bi = hB, 0i
⇒ c2 = 0.X
Thus { A, B } is a linearly independent set.
(2.3) Apply the Gram-Schmidt process to the following subset of M22 : (12)

1 1 1 1 0 1
, ,
0 2 0 0 0 1
Let
1 1 1 1 0 1
u1 := , u2 := , u3 := .
0 2 0 0 0 1
187

1 1
v1 := u1 =
0 2
X
T !
2 0 1 1 1 1
hv1 , v1 i = tr = 8X
0 1 0 2 0 2
T !
2 0 1 1 1 1
hu2 , v1 i = tr = 4X
0 1 0 0 0 2
hu2 , v1 i
v2 := u2 − v1 (x)X
hv1 , v1 i

1 1 4 1 1 1 1 1
= − = X
0 0 8 0 2 2 0 −2
T !
1 2 0 1 1 1 1
hv2 , v2 i = tr = 2X
4 0 1 0 −2 0 −2
T !
2 0 0 1 1 1
hu3 , v1 i = tr = 4X
0 1 0 1 0 2
T !
1 2 0 0 1 1 1
hu3 , v2 i = tr = 0X
2 0 1 0 1 0 −2
hu3 , v1 i hu3 , v2 i
v3 := u3 − v1 − v2 X
hv1 , v1 i hv2 , v2 i

0 1 4 1 1 0 1 1 1
= − − ·
0 1 8 0 2 2 2 0 −2

1 −1 1
= .X
2 0 0

1 1 1 1 −1 1
, , .X 2
0 2 0 −2 0 0
(2.4) Let V be a vector space with zero vector 0 and let h·, ·i denote an inner product on V . (4)
Prove that h0, vi = 0 for all v ∈ V .
Since 0 = 0 · 0X2 (Theorem VZ) we have h0, vi = h0 · 0, vi = 0h0, vi = 0X2 by IP2.
Alternative:
Since 0 = 0 + 0 (VS4) we have h0, vi = h0 + 0, vi = h0, vi + h0, vi by IP3 and IP1, so
that h0, vi = 0. Alternative:
p
Since the Cauchy-Schwarz inequality yields |h0, vi| ≤ h0, 0ihv, vi = 0 by IP4b, it
follows that |h0, vi| = 0. Thus h0, vi = 0.
188
MAT2611/101/3/2017

1 1 0
A = 0 1 0 .
1 1 0

Row reduction of A yields
   
1 1 0 1 1 0
0 1 0 → 0 1 0 (R3 ← R3 − R1 )
1 1 0 0 0 0
which is in upper triangular form, with two nonzero rows. Hence the rank is 2, and the
nullity is 3 − 2 = 1.X2
λ(λ − 1)2 = 0.
    
1 0 0 1 1 0 λ − 1 −1 0
det λ 0 1 0 − 0 1 0
     X= 0 λ−1 0
0 0 1 1 1 0 −1 −1 λ
= (λ − 1)2 λ = 0X2 .

From the characteristic equation we obtain the eigenvalues 0, and 1 (twice)X2 . For the
eigenspace corresponding to the eigenvalue 0 we solve
    
−1 −1 0 x 0
 0 −1 0 y  = 0 X2
−1 −1 0 z 0
for x, y, z ∈ R. Clearly x = −y = 0. We find the 1-dimensional eigenspace

      
 0   0 
0 : z ∈ R = z 0 : z ∈ R .X2
z 1
   
Thus a basis is given by  

 0 
0 .X2
1
 
189
For the eigenspace corresponding to the eigenvalue 1 we solve
    
0 −1 0 x 0
0 0 0 y  = 0 X2
−1 −1 1 z 0
for x, y, z ∈ R. Obviously y = 0 and x = z. The corresponding eigenspace is
      
 x   1 
 0  : x ∈ R = x 0 : x ∈ R .X2
x 1
   

 1 
0 .X2
1
 
(3.4) For each eigenvalue, determine the algebraic and geometric multiplicity. Is A diagonal- (5)
izable?
The algebraic multiplicity of λ = 1 is 2,Xand the geometric multiplicity is 1.XThus A
is not diagonalizable (Theorem DM).XThe algebraic multiplicity of λ = 0 is 1,Xand
the geometric multiplicity is 1.X
(3.5) Prove or disprove: (2)

If B is a 2 × 2 matrix, then B is diagonalizable if and only if B 2 is diagonalizable.

0 1
The statement is false, for example the matrix B = is not diagonalizable, but
0 0
0 0
2
B = is diagonalizable (and diagonal).X2
0 0
(3.6) Let B be an n × n matrix. Prove that B + B T is diagonalizable. (2)

Since (B + B T )T = B T + (B T )T = B T + B = B + B T , B + B T is symmetricXand
consequently diagonalizable (Theorem DS).X

3 x y
Let T : R → M22 be defined by T (x, y, z) = .
z x

3
Let k ∈ R and (x1 , y1 , z1 ), (x2 , y2 , z2 ) ∈ R .

x1 + x 2 y 1 + y 2
• T ((x1 , y1 , z1 ) + (x2 , y2 , z2 )) = T (x1 + x2 , y1 + y2 , z1 + z2 ) =
z1 + z2 x1 + x2
x y x y
= 1 1 + 2 2 = T (x1 , y1 , z1 )+T (x2 , y2 , z2 ).X2
z1 x1 z2 x2

kx1 ky1 x1 y 1
• T (k·(x1 , y1 , z1 )) = T (kx1 , ky1 , kz1 ) = =k = kT (x1 , y1 , z1 ).X2
kz1 kx1 z1 x1
190
MAT2611/101/3/2017
(4.2) Find the matrix representation [T ]B2 ,B1 of T relative to the basis (8)
B1 = { (1, 0, 1), (0, 1, 0), (1, 0, −1) }
in R3 and the basis

1 0 1 0 0 1 0 1
B2 = , , ,
0 1 0 −1 1 0 −1 0
in M22 , ordered from left to right.
From

1 0 1 0 1 0 1 0 1 1 0 1
T (1, 0, 1) = =1 +0
0 −1
+ − X2
1 1 0 1 2 1 0 2 −1 0

0 1 1 0 1 0 1 0 1 1 0 1
T (0, 1, 0) = =0 +0
0 −1
+ + X2
0 0 0 1 2 1 0 2 −1 0

1 0 1 0 1 0 1 0 1 1 0 1
T (1, 0, −1) =
−1 1
=1 +0
0 −1
− + X2
0 1 2 1 0 2 −1 0
the coefficients of the basis elements in each equation provide the columns of the matrix
representation:  
2 0 2
1 0 0 0 .X 2
2  1 1 −1
−1 1 1
(4.3) Determine the range R(T ) of T . Is T onto? In other words, is it true that R(T ) = M22 ? (4)
The range of T is
R(T ) = { T (x, y, z) : x, y, z ∈ R }

x y
= : x, y, z ∈ R .X2
z x

1 0 1 0
Since ∈ M22 but / R(T ), T is not onto.X2
∈
0 0 0 0

3 0 0
ker(T ) = (x, y, z) ∈ R : T (x, y, z) =
0 0

3 x y 0 0
= (x, y, z) ∈ R : =
z x 0 0
= { (0, 0, 0) } .X2
We have a zero-dimensional space and the nullity of T is 0. X2

Yes, since nullity(T ) = 0 (or equivalently ker(T ) = { (0, 0, 0) }, Theorem TO).X2
191
Question paper
This question is a multiple choice question and should be answered in the green answer book.
Any rough work should be clearly marked and appear on the last pages of the answer book.

X := { (x, y) : x, y ∈ R }
· : R × X → X, k · a ≡ k · (x, y) := (kx + k − 1, ky),

+ : X × X → X, a + b ≡ (x, y) + (α, β) := (x + α + 1, y + β).
The set X with these definitions of · and + forms a vector space. Which one of the
following statements is true in X ?
1. −(1, 1) = (−3, −1)
2. −(1, 1) = (−2, −1)
3. −(1, 1) = (−1, −1)
4. −(1, 1) = (0, −1)
(1.2) Which of the following are subspaces of P1 with the usual operations ? (2)
A. span { 1 + x }
B. { ax : a ∈ R }
C. { 1 + ax : a ∈ R }
D. { (1 + a)x : a ∈ R }
2. Only A, B and D.
3. Only A, B and C.
4. Only B, C and D.
192
MAT2611/101/3/2017
A. span { (1, 0, 1), (1, 0, 2) } in R3
B. { (1, 0, 1), (1, 0, 2) } in R3

1 1 1 2
C. , in M22
1 1 1 1
1. Only A.
2. Only A and B.
3. Only B.
4. Only B and C.
(1.4) Which of the following sets are a basis for the following vector subspace of P2 : (2)
X = { p(x) ∈ P2 : p(1) = 0 } .
A. { 1 − 2x + x2 , 2 − 3x + x2 }
B. { 1 − x }
C. { 1 − 2x + x2 }
1. Only A.
2. Only B.
3. Only C.
4. Only B and C.
193
A. dim (span { (1, 1, 1), (1, 1, 0) }) = 2 in R3
B. dim (span { (0, 0, 0), (1, 1, 1), (1, 1, 0) }) = 3 in R3

1 1
C. dim span = 2 in M22
1 −1
2. Only A.
3. Only A and B.
4. Only A and C.

(1.6) Which of the following sets are a basis for the row space of 1 −1 −1 1 ? (2)

A. −1 1 1 −1

B. 1 −1

C. 1 −1 , −1 1
1. Only A.
2. Only B.
3. Only C.
4. Only A and C.

(1.7) Which of the following sets are a basis for the column space of 1 −1 −1 1 ? (2)

A. 1 , −1

B. −1

C. 1 −1

D. 1 −1 , −1 1
1. Only A.
2. Only B.
3. Only C.
4. Only D.
194
MAT2611/101/3/2017

(1.8) Which one of the following statements is true for the matrix A = 1 −1 −1 1 ? (2)
Consider the vector space P3 .

hp(x), q(x)i := p0 q0 + 2p1 q1 + 2p2 q2 + p3 q3 ,
where
p(x) = p0 + p1 x + p2 x2 + p3 x3 and q(x) = q0 + q1 x + q2 x2 + q3 x3 ,
is an inner product on P3 .
(2.2) Are the vectors (6)

1 + x2 + x3 , −1 − x2 + x3 , −1 + x − x2 + x3
linearly independent?
(2.3) Apply the Gram-Schmidt process to the following subset of P3 : (12)
1 + x2 + x3 , −1 − x2 + x3 , −1 + x − x2 + x3


1 1 1
A =  3 0 0 .
−1 2 2
λ2 (λ − 3) = 0.
195
(3.4) Is A diagonalizable? Motivate your answer. (2)
(3.5) Let B be an n × n matrix and B T be the transpose of B. Prove that: (5)
B is diagonalizable if and only if B T is diagonalizable.
Hint: recall that (P −1 )T = (P T )−1 for any invertible matrix P .
Let T : M22 → R2 be defined by

a b
T = (a, b) + (c, d)
c d
where a, b, c, d ∈ R.
(4.2) Find the matrix representation [T ]B 0 ,B of T relative to the basis (10)

1 1 0 0 1 −1 0 0
B= , , ,
0 0 1 1 0 0 1 −1
in M22 , and the basis

B 0 = { (1, 1), (1, −1) }
in R2 , ordered from left to right.
(4.3) Determine the range R(T ) of T . Is T onto? In other words, is it true that R(T ) = R2 ? (4)
196
MAT2611/101/3/2017
Solution
This question is a multiple choice question and should be answered in the green answer book.
Any rough work should be clearly marked and appear on the last pages of the answer book.

X := { (x, y) : x, y ∈ R }
· : R × X → X, k · a ≡ k · (x, y) := (kx + k − 1, ky),

+ : X × X → X, a + b ≡ (x, y) + (α, β) := (x + α + 1, y + β).
1. −(1, 1) = (−3, −1)
2. −(1, 1) = (−2, −1)
3. −(1, 1) = (−1, −1)
4. −(1, 1) = (0, −1)

Answer: 1
(1.2) Which of the following are subspaces of P1 with the usual operations ? (2)
A. span { 1 + x }
B. { ax : a ∈ R }
C. { 1 + ax : a ∈ R }
D. { (1 + a)x : a ∈ R }
2. Only A, B and D.
3. Only A, B and C.
4. Only B, C and D.

Answer: 2
197
A. span { (1, 0, 1), (1, 0, 2) } in R3
B. { (1, 0, 1), (1, 0, 2) } in R3

1 1 1 2
C. , in M22
1 1 1 1
1. Only A.
2. Only A and B.
3. Only B.
4. Only B and C.

Answer: 4
X = { p(x) ∈ P2 : p(1) = 0 } .
A. { 1 − 2x + x2 , 2 − 3x + x2 }
B. { 1 − x }
C. { 1 − 2x + x2 }
1. Only A.
2. Only B.
3. Only C.
4. Only B and C.

Answer: 1
198
MAT2611/101/3/2017
A. dim (span { (1, 1, 1), (1, 1, 0) }) = 2 in R3
B. dim (span { (0, 0, 0), (1, 1, 1), (1, 1, 0) }) = 3 in R3

1 1
C. dim span = 2 in M22
1 −1
2. Only A.
3. Only A and B.
4. Only A and C.

Answer: 2

(1.6) Which of the following sets are a basis for the row space of 1 −1 −1 1 ? (2)

A. −1 1 1 −1

B. 1 −1

C. 1 −1 , −1 1
1. Only A.
2. Only B.
3. Only C.
4. Only A and C.

Answer: 1
199

(1.7) Which of the following sets are a basis for the column space of 1 −1 −1 1 ? (2)

A. 1 , −1

B. −1

C. 1 −1

D. 1 −1 , −1 1
1. Only A.
2. Only B.
3. Only C.
4. Only D.

Answer: 2

(1.8) Which one of the following statements is true for the matrix A = 1 −1 −1 1 ? (2)

Answer: 2
200
MAT2611/101/3/2017

hp(x), q(x)i := p0 q0 + 2p1 q1 + 2p2 q2 + p3 q3 ,
where
p(x) = p0 + p1 x + p2 x2 + p3 x3 and q(x) = q0 + q1 x + q2 x2 + q3 x3 ,
is an inner product on P3 .
p(x) = p0 + p1 x + p2 x2 + p3 x3 , q(x) = q0 + q1 x + q2 x2 + q3 x3 , r(x) = r0 + r1 x + r2 x2 + r3 x3
1. hp(x), q(x)i = p0 q0 +2p1 q1 +2p2 q2 +p3 q3 = q0 p0 +2q1 p1 +2q2 p2 +q3 p3 = hq(x), p(x)i
X2
2. hp(x) + r(x), q(x)i = (p0 + r0 )q0 + 2(p1 + r1 )q1 + 2(p2 + r2 )q2 + (p3 + r3 )q3
= p0 q0 + r0 q0 + 2p1 q1 + 2r1 q1 + 2p2 q2 + 2r2 q2 + p3 q3 + r3 q3
= p0 q0 + 2p1 q1 + 2p2 q2 + p3 q3 + r0 q0 + 2r1 q1 + 2r2 q2 + r3 q3
= hp(x), q(x)i + hr(x), q(x)i X4
3. hkp(x), q(x)i = (kp0 )q0 + 2(kp1 )q1 + 2(kp2 )q2 + (kp3 )q3 = k(p0 q0 + 2p1 q1 + 2p2 q2 +
p3 q3 ) = khp(x), q(x)iX2
4. hp(x), p(x)i = p20 +2p21 +2p22 +p23 ≥ 0X2 so that hp(x), p(x)i ≥ 0 and hp(x), p(x)i =
0 if and only if p0 = p1 = p2 = p3 = 0 (since p20 , p21 , p22 , p23 ≥ 0)X2 , i.e. p(x) = 0.

1 + x2 + x3 , −1 − x2 + x3 , −1 + x − x2 + x3
Consider the equation
a(1 + x2 + x3 ) + b(−1 − x2 + x3 ) + c(−1 + x − x2 + x3 ) = 0X2
where a, b, c ∈ R. Thus we have the equations
a−b−c=0
c=0
a−b−c=0
a+b+c=0
Adding the fourth equation to the third provides a = 0. The second equation provides
c = 0. Inserting the solutions for a and c into the first equation yields b = 0. This is
the only solution. Thus these vectors are linearly independent.X4
201
(2.3) Apply the Gram-Schmidt process to the following subset of P3 : (12)
1 + x2 + x3 , −1 − x2 + x3 , −1 + x − x2 + x3

Let
u1 (x) := 1 + x2 + x3 , u2 (x) := −1 − x2 + x3 , u3 (x) := −1 + x − x2 + x3 .
v1 (x) := u1 = 1 + x2 + x3 X
hv1 (x), v1 (x)i = 12 + 2 · 02 + 2 · 12 + 12 = 4X
hu2 (x), v1 (x)i = (−1) · 1 + 2 · 0 · 0 + 2 · (−1) · 1 + 1 · 1 = −2X
hu2 (x), v1 (x)i
v2 (x) := u2 (x) − v1 (x)X
hv1 (x), v1 (x)i
−2 1
= (−1 − x2 + x3 ) − (1 + x2 + x3 ) = (−1 − x2 + 3x3 )X
4 2
1
hv2 (x), v2 (x)i = ((−1)2 + 2 · 02 + 2(−1)2 + 32 ) = 3X
4
hu3 (x), v1 (x)i = (−1) · 1 + 2 · 1 · 0 + 2 · (−1) · 1 + 1 · 1 = −2X

1 1 3
hu3 (x), v2 (x)i = (−1) · − + 2 · 1 · 0 + 2 · (−1) · − + 1 · = 3X
2 2 2
hu3 (x), v1 (x)i hu3 (x), v2 (x)i
v3 (x) := u3 (x) − v1 − v2 X
hv1 (x), v1 (x)i hv2 (x), v2 (x)i
−2 3 1
= (−1 + x − x2 + x3 ) − (1 + x2 + x3 ) − · (−1 − x2 + 3x3 )
4 3 2
= x.X

3 1
1 + x + x , (−1 − x + 3x ), x .X2
2 2 3
2

1 1 1
A =  3 0 0 .
−1 2 2
202
MAT2611/101/3/2017

   
1 1 1 1 1 1
3 0 0 →  0 −3 −3 (R2 ← R2 − 3R1 )
−1 2 2 −1 2 2
 
1 1 1
→ 0 −3 −3 (R3 ← R3 + R1 )
0 3 3
 
1 1 1
→ 0 −3 −3 (R3 ← R2 + R3 )
0 0 0
which is in upper triangular form, with two nonzero rows. Hence the rank is 2, and the
nullity is 3 − 2 = 1.X2
λ2 (λ − 3) = 0.
    
1 0 0 1 1 1 λ − 1 −1 −1
det λ 0 1 0 − 3 0 0
      X = −3 λ 0
0 0 1 −1 2 2 1 −2 λ − 2
= λ(λ − 1)(λ − 2) − 6 − 3(λ − 2) + λ = λ3 − 3λ2
= λ2 (λ − 3) = 0X2 .

From the characteristic equation we obtain the eigenvalues 0 (twice), and 3X2 . For the
    
−1 −1 −1 x 0
−3 0 0   y = 0 X2
 
1 −2 −2 z 0
203
for x, y, z ∈ R. Row reduction yields
   
−1 −1 −1 1 1 1
−3 0 0  → −3 0 0 (R1 ← −R1 )
1 −2 −2 1 −2 −2
 
1 1 1
→ 0 3 3 (R2 ← R2 + 3R1 )
1 −2 −2
 
1 1 1
→ 0 3 3 (R3 ← R3 − R1 )
0 −3 −3
 
1 1 1
→ 0 3 3 (R3 ← R2 + R3 )
0 0 0
 
1 1 1
→ 0 1 1 (R2 ← R2 /3)
0 0 0
 
1 0 0
→ 0 1 1 (R1 ← R1 − R2 )
0 0 0
so that x = 0 and y = −z. We find the 1-dimensional eigenspace

      
 0   0 
 z  : z ∈ R = z  1  : z ∈ R .X 4
−z −1
   

 0 
 1  .X 2
−1
 

    
2 −1 −1 x 0
−3 3 0  y  = 0 X2
1 −2 1 z 0
204
MAT2611/101/3/2017

   
2 −1 −1 1 −2 1
−3 3 0  → −3 3 0 (R1 ↔ R3 )
1 −2 1 2 −1 −1
 
1 −2 1
→ 0 −3 3  (R2 ← R2 + 3R1 )
2 −1 −1
 
1 −2 1
→ 0 −3 3  (R3 ← R2 − 2R1 )
0 3 −3
 
1 −2 1
→ 0 −3 3 (R3 ← R3 + R2 )
0 0 0
 
1 −2 1
→ 0 1 −1 (R2 ← −R2 /3)
0 0 0
 
1 0 −1
→ 0 1 −1 (R1 ← R1 + 2R2 )
0 0 0
so that x = y = z. The corresponding eigenspace is

      
 z   1 
z  : z ∈ R = z 1 : z ∈ R .X4
z 1
   

 1 
1 .X2
1
 

No, because the geometric and algebraic multiplicities are not equal for the eigenvalue
0 (i.e. the algebraic multiplicity is 2 while the geometric multiplicity is 1).X2
(3.5) Let B be an n × n matrix and B T be the transpose of B. Prove that: (5)

B is diagonalizable if and only if B T is diagonalizable.
Hint: recall that (P −1 )T = (P T )−1 for any invertible matrix P .
Assume B is diagonalizable, then there exists an invertible n × n matrix P such that
P −1 BP is diagonal.XIt follows that
(P −1 BP )T = P T B T (P −1 )T
is diagonal (since the transpose of a diagonal matrix is diagonal).XLet Q = (P −1 )T

so that Q−1 = P T . Obviously Q is invertible and Q−1 B T Q is diagonal.XHence B T is
205
diagonalizable.X
Similarly, if B T is diagonalizable then there exists invertible Q such that Q−1 BQ is

diagonal. Let P = (Q−1 )T . Then P −1 BP is diagonal.X
Let T : M22 → R2 be defined by

a b
T = (a, b) + (c, d)
c d

2
Let k, a, b, c, d, α, β, γ, δ ∈ R. Using the definition of R , M22 and T we find
•

a b ka kb
T k =T = (ka, kb) + (kc, kd) = k(a, b) + k(c, d)
c d kc kd

a b
= k((a, b) + (c, d)) = kT .X 2
c d
•

a b α β a+α b+β
T + =T = (a + α, b + β) + (c + γ, d + δ)
c d γ δ c+γ d+δ

a b α β
= ((a, b) + (c, d)) + ((α, β) + (γ, δ)) = T T +T .X 2
c d γ δ

1 1 0 0 1 −1 0 0
B= , , ,
0 0 1 1 0 0 1 −1
B 0 = { (1, 1), (1, −1) }
in R2 , ordered from left to right.
From

1 1
T = (1, 1) + (0, 0) = (1, 1) = 1 · (1, 1) + 0 · (1, −1)X2
0 0

0 0
T = (0, 0) + (1, 1) = (1, 1) = 1 · (1, 1) + 0 · (1, −1)X2
1 1

1 −1
T = (1, −1) + (0, 0) = (1, −1) = 0 · (1, 1) + 1 · (1, −1)X2
0 0

0 0
T = (0, 0) + (1, −1) = (1, −1) = 0 · (1, 1) + 1 · (1, −1)X2
1 −1
206
MAT2611/101/3/2017
representation:
1 1 0 0
.X2
0 0 1 1
(4.3) Determine the range R(T ) of T . Is T onto? In other words, is it true that R(T ) = R2 ? (4)
T is onto since

a b
R(T ) = T : a, b, c, d ∈ R X2
c d
= { (a, b) + (c, d) : a, b, c, d ∈ R }
= { (α, β) : α, β ∈ R } (setting α = a + c and β = b + d)
= R2 .X2

a b a b
ker(T ) = : a, b, c, d ∈ R, T = (0, 0)
c d c d

a b
= : a, b, c, d ∈ R, (a, b) + (c, d) = (a + c, b + d) = 0
c d

a b
= : a, b ∈ R, c = −a, d = −b
c d

a b
= : a, b ∈ R
−a −b

1 0 0 1
= a +b : a, b ∈ R .X2
−1 0 0 −1
Since
1 0 0 1
,
−1 0 0 −1
is a linearly independent set, we have a two-dimensional space and the nullity of T is 2.
X2
No, since (for example)

1 0 0 0
T =T = (1, 0).X2
0 0 1 0
207
Question paper
This question is a multiple choice question and should be filled in on the multiple choice answer
sheet (mark reading sheet).

X := { (x, y) : x, y ∈ R }
· : R × X → X, k · a ≡ k · (x, y) := (kx − k + 1, ky),

+ : X × X → X, a + b ≡ (x, y) + (α, β) := (x + α − 1, y + β).
1. −(0, 0) = (1, 0)
2. −(0, 0) = (1, 1)
3. −(0, 0) = (0, 1)
4. −(0, 0) = (2, 0)
(1.2) Which of the following are subspaces of R2 with the usual operations ? (2)
A. span { (2, 3) }
B. { (x, 1) : x ∈ R }
C. { (0, x) : x ∈ R, x ≥ 0 }
D. { (0, x − 1) : x ∈ R }
1. Only A.
2. Only A and D.
3. Only C.
4. Only C and D.
208
MAT2611/101/3/2017
A. span { (2, 3) } in R2
B. { (1, 1), (−1, 1) } in R2
C. { (2, 4), (1, −1), (1, 1) } in R2
D. { 1 + x, 1 − x } in P1
1. Only A.
2. Only B.
3. Only B and C.
4. Only B and D.
X = { p(x) ∈ P2 : p(1) = 0 } .
A. { 1, x, x2 }
B. { 1 − x, 1 − x2 }
C. { 1, 1 − x, 1 − x2 }
D. { 1 − x, 1 − x2 , 3 − 2x − x2 }
1. Only A.
2. Only B.
3. Only A and C.
4. A, B, C and D.
209
A. dim(span { (1, 1, 1), (1, 1, −1) }) = 2 in R3
B. dim(span { (0, 0, 0), (1, 1, 1) }) = 2 in R3
C. dim(span { (1, 1, 1), (1, −1, 1), (1, 1, −1) }) = 2 in R3
D. dim(span { (1, 1, 1), (1, −1, 1), (1, 1, −1) }) = 3 in R3
2. Only A and B.
3. Only A and C.
4. Only A and D.

3 −1 2
(1.6) Which of the following sets are a basis for the row space of ? (2)
3 2 −1

A. 0 −3 3 , 3 −1 2

B. 0 −3 3 , 3 0 1

C. 3 −1 2 , 3 2 −1
1. Only A.
2. Only B.
3. Only A and B.
4. A, B and C.
210
MAT2611/101/3/2017
(1.7) Which
of the
following sets are contained in (i.e. subset of) the column space of (2)
3 −1 2
?
3 2 −1
n T T o
A. 1 1 , −1 2
n T T o
B. −1 2 , 2 1
n T T T o
C. 1 0 , 0 1 , 1 1

D. 0 3 −3
1. Only D.
2. Only A, B and C.
3. Only A and B.
4. Only A and C.

3 −1 2
(1.8) Which of the following sets are a basis for the null space of ? (2)
3 2 −1
n T T o
A. 1 1 , −1 2
n T o
B. −1 1 1
n T o
C. −1 3 3
1. Only B and C.
2. Only B.
3. Only C.
4. Only A.
211

   
x1 y1
x2  y2  4
hx, yi := 2x1 y1 + 2x2 y2 + x3 y3 + x4 y4 , x3  , y = y3  ∈ R
x=   
x4 y4

     
1 1 1
0 0 1
 ,  ,  
1 −1 −1
1 1 1
(2.3) Apply the Gram-Schmidt process to the following subset of R4 : (12)

     

 1 1 1 
     
0 ,  0  ,  1 

 1 −1 −1 
 
1 1 1
 
to find an orthogonal basis with respect to the inner product defined in 2.1 for the

1 0 0
A =  21 1
2
1
2
.
− 21 1
2
1
2
(3.1) Determine the rank of A. (2)
λ(λ − 1)2 = 0.
212
MAT2611/101/3/2017
(3.5) Is the matrix A − I3 diagonalizable? Motivate your answer. (Here I3 is the 3 × 3 identity (2)
matrix).
Let T : M22 → P2 be defined by

a b b−c
T =a+ x + dx2
c d 2
where a, b, c ∈ R.

1 0 0 0 0 1 0 1
B= , , ,
0 0 0 1 1 0 −1 0

B 0 = { 1 + x, 1 − x, x2 }
in P2 , ordered from left to right.
(4.3) Determine the range R(T ) of T . Is T onto? In other words, is it true that R(T ) = P2 ? (4)
213
Solution
This question is a multiple choice question and should be filled in on the multiple choice answer
sheet (mark reading sheet).

X := { (x, y) : x, y ∈ R }
· : R × X → X, k · a ≡ k · (x, y) := (kx − k + 1, ky),

+ : X × X → X, a + b ≡ (x, y) + (α, β) := (x + α − 1, y + β).
1. −(0, 0) = (1, 0)
2. −(0, 0) = (1, 1)
3. −(0, 0) = (0, 1)
4. −(0, 0) = (2, 0)

Answer: 4
(1.2) Which of the following are subspaces of R2 with the usual operations ? (2)
A. span { (2, 3) }
B. { (x, 1) : x ∈ R }
C. { (0, x) : x ∈ R, x ≥ 0 }
D. { (0, x − 1) : x ∈ R }
1. Only A.
2. Only A and D.
3. Only C.
4. Only C and D.

Answer: 2
214
MAT2611/101/3/2017
A. span { (2, 3) } in R2
B. { (1, 1), (−1, 1) } in R2
C. { (2, 4), (1, −1), (1, 1) } in R2
D. { 1 + x, 1 − x } in P1
1. Only A.
2. Only B.
3. Only B and C.
4. Only B and D.

Answer: 4
X = { p(x) ∈ P2 : p(1) = 0 } .
A. { 1, x, x2 }
B. { 1 − x, 1 − x2 }
C. { 1, 1 − x, 1 − x2 }
D. { 1 − x, 1 − x2 , 3 − 2x − x2 }
1. Only A.
2. Only B.
3. Only A and C.
4. A, B, C and D.

Answer: 2
215
A. dim(span { (1, 1, 1), (1, 1, −1) }) = 2 in R3
B. dim(span { (0, 0, 0), (1, 1, 1) }) = 2 in R3
C. dim(span { (1, 1, 1), (1, −1, 1), (1, 1, −1) }) = 2 in R3
D. dim(span { (1, 1, 1), (1, −1, 1), (1, 1, −1) }) = 3 in R3
2. Only A and B.
3. Only A and C.
4. Only A and D.

Answer: 4

3 −1 2
(1.6) Which of the following sets are a basis for the row space of ? (2)
3 2 −1

A. 0 −3 3 , 3 −1 2

B. 0 −3 3 , 3 0 1

C. 3 −1 2 , 3 2 −1
1. Only A.
2. Only B.
3. Only A and B.
4. A, B and C.

Answer: 4
216
MAT2611/101/3/2017
(1.7) Which
of the
following sets are contained in (i.e. subset of) the column space of (2)
3 −1 2
?
3 2 −1
n T T o
A. 1 1 , −1 2
n T T o
B. −1 2 , 2 1
n T T T o
C. 1 0 , 0 1 , 1 1

D. 0 3 −3
1. Only D.
2. Only A, B and C.
3. Only A and B.
4. Only A and C.

Answer: 2

3 −1 2
(1.8) Which of the following sets are a basis for the null space of ? (2)
3 2 −1
n T T o
A. 1 1 , −1 2
n T o
B. −1 1 1
n T o
C. −1 3 3
1. Only B and C.
2. Only B.
3. Only C.
4. Only A.

Answer: 3
217

   
x1 y1
x2  y2  4
hx, yi := 2x1 y1 + 2x2 y2 + x3 y3 + x4 y4 , x3  , y = y3  ∈ R
x=   
x4 y4
     
x1 y1 z1
x2  y2  z2  4
x3  , y = y3  , z = z3  ∈ R
x=     
x4 y4 z4
1. hx, yi = 2x1 y1 + 2x2 y2 + x3 y3 + x4 y4 = 2y1 x1 + 2y2 x2 + y3 x3 + y4 x4 = hy, xi X2
2. hx + z, yi = 2(x1 + z1 )y1 + 2(x2 + z2 )y2 + (x3 + z3 )y3 + (x4 + z4 )y4
= 2x1 y1 + 2z1 y1 + 2x2 y2 + 2z2 y2 + x3 y3 + z3 y3 + x4 y4 + z4 y4
= 2x1 y1 +2x2 y2 +x3 y3 +x4 y4 +2z1 y1 +2z2 y2 +z3 y3 +z4 y4 = hx, yi+hz, yi
X4
3. hkx, yi = 2(kx1 )y1 +2(kx2 )y2 +(kx3 )y3 +(kx4 )y4 = k(2x1 y1 +2x2 y2 +x3 y3 +x4 y4 ) =
khx, yiX2
4. hx, xi = 2x21 + 2x22 + x23 + x24 ≥ 0X2 so that hx, xi ≥ 0 and hx, xi = 0 if and only
if x1 = x2 = x3 = x4 = 0 (since x21 , x22 , x23 , x24 ≥ 0)X2 , i.e. x = 0.

     
1 1 1
0 0 1
 ,  ,  
1 −1 −1
1 1 1
     
1 1 1
0 0 1 2
1 + b −1 + c −1 = 0X
a     
1 1 1
a+b+c=0
c=0
a−b−c=0
a+b+c=0
218
MAT2611/101/3/2017
Adding the fourth equation to the third provides a = 0. The second equation provides

     

 1 1 1 
     
0 0
 ,  ,  1 

 1 −1 −1 
 
1 1 1
 
Let      
1 1 1
0 0 1
u1 := 
1 , u2 := −1 , u3 := −1 .
    
1 1 1
 
1
0
1 X
v1 := u1 =  
1
hv1 , v1 i = 2 · 12 + 2 · 02 + 12 + 12 = 4X
hu2 , v1 i = 2 · 1 · 1 + 2 · 0 · 0 + 1 · (−1) + 1 · 1 = 2X
     
1 1 1
hu2 , v1 i  0  2 0 1  0 
   
v2 := u2 − v1 X = −1 − 4 1 = 2 −3 X

hv1 , v1 i
1 1 1
1
hv2 , v2 i = (2 · 12 + 2 · 02 + (−3)2 + 12 ) = 3X
4
hu3 , v1 i = 2 · 1 · 1 + 2 · 1 · 0 + (−1) · 1 + 1 · 1 = 2X

1 3 1
hu3 , v2 i = 2 · 1 · + 2 · 1 · 0 + (−1) · − + 1 · = 3X
2 2 2
hu3 , v1 i hu3 , v2 i
v3 := u3 − v1 − v2 X
hv1 , v1 i hv2 , v2 i
     
1 1 1
 1  2 0 3 1  0 
=−1 − 4 1 − 3 · 2 −3
    
1 1 1
 
0
1
=0 .X

219
     

 1 1 0 
  1  
0
 ,  0 ,
1  2
  .X
 1 2 −3 0 
 
1 1 0
 

1 0 0
A =  21 1
2
1
2
.
− 21 1
2
1
2
(3.1) Determine the rank of A. (2)

   
1 0 0 1 0 0
 1 1 1  → 0 1 1  (R2 ← R2 − R1 /2, R3 ← R3 + R1 /2)
2 2 2 2 2
− 12 12 21 0 12 21
 
1 0 0
→ 0 21 21  (R3 ← R3 − R2 )
0 0 0
which is in upper triangular form, with two nonzero rows. Hence the rank is 2.X2
λ(λ − 1)2 = 0.
    
1 0 0 1 0 0 λ−1 0 0
1 1 1  2
det λ 0 1 0 − 2 2 2
    X = − 2 λ − 2 − 12
1 1
1 1 1 1
0 0 1 −2 2 2 2
− 21 λ − 21
λ − 21 − 12
= (λ − 1)
− 12 λ − 12
2 !
1 1
= (λ − 1) λ− − = (λ − 1)(λ − 1)λ = 0X4 .
2 4
220
MAT2611/101/3/2017

From the characteristic equation we obtain the eigenvalues 0, and 1X2 (twice). For the
    
−1 0 0 x 0
− 1 − 1 − 1  y  = 0 X2
2 2 2
1 1 1
2
− 2
− 2
z 0

   
−1 0 0 −1 0 0
− 1 − 1 − 1  →  0 − 1 − 1  (R2 ← R2 − R1 /2, R3 ← R3 + R1 /2)
2 2 2 2 2
1 1 1 1 1
2
− 2
− 2
0 − 2
− 2
 
−1 0 0
→  0 − 21 − 12  (R3 ← R3 − R2 )
0 0 0
 
1 0 0
→ 0 1 1
 (R1 ← −R1 , R2 ← −2R2 )
0 0 0
so that x = 0 and y = −z. We find the 1-dimensional eigenspace

      
 0   0 
−z  : z ∈ R = z −1 : z ∈ R .X4
z 1
   

 0 
−1 .X2
1
 

    
0 0 0 x 0
− 1 1 − 1  y  = 0 X2
2 2 2
1
2
− 12 21 z 0

 1
− 12 12
  
0 0 0 2
− 1 1 − 1  → − 1 1 − 1  (R1 ↔ R3 )
2 2 2 2 2 2
1
2
− 12 12 0 0 0
1 1 1

2
− 2 2
→ 0 0 0 (R2 ← R2 + R1 )
0 0 0
 
1 −1 1
→ 0 0 0 (R1 ← 2R1 )
0 0 0
221
Thus we find the 2-dimensional eigenspace
        
 y−z   1 −1 
 y  : y, z ∈ R = y 1 + z  0  : x, y ∈ R .X4
z 0 1
   
Thus a basis is given by    

 1 −1 
1 ,  0  .X2
0 1
 

Yes, because the geometric and algebraic multiplicities are equal for each eigenvalueX2
(i.e. 1 for λ = 0 and 2 for λ = 1).
(3.5) Is the matrix A − I3 diagonalizable? Motivate your answer. (Here I3 is the 3 × 3 identity (2)
matrix).
Yes. Since the eigenvalues of A − I3 are 0 − 1 = −1 and 1 − 1 = 0 with algebraic
multiplicities 1 and 2 respectively. The eigenspaces of A are also eigenspaces of A − I3 ,
i.e. the eigenspace E0 of A (with dimension 1) is the eigenspace of A − I3 corresponding
to the eigenvalue -1, and similarly E1 (with dimension 2) to the eigenvalue 0. Thus the
geometric and algebraic multiplicities are equal.X2
Let T : M22 → P2 be defined by

a b b−c
T =a+ x + dx2
c d 2

Let k, a, b, c, d, α, β, γ, δ ∈ R. Using the definition of M22 and T we find
•

a b ka kb
T k =T
c d kc kd
kb − kc
= ka + x + kdx2
2
b−c 2
=k a+ x + dx
2

a b
= kT .X 2
c d
222
MAT2611/101/3/2017
•

a b α β a+α b+β
T + =T
c d γ δ c+γ d+δ
b + β − (c + γ)
= (a + α) + x + (d + δ)x2
2
b−c+β−γ
= (a + α) + x + (d + δ)x2
2
b−c β−γ
= (a + x + dx2 ) + (α + x + δx2 )
2 2
a b α β
=T +T .X 2
c d γ δ

1 0 0 0 0 1 0 1
B= , , ,
0 0 0 1 1 0 −1 0
B 0 = { 1 + x, 1 − x, x2 }
in P2 , ordered from left to right.
From

1 0 0−0 1 1
T =1+ x + 0x2 = · (1 + x) + · (1 − x) + 0 · x2 X2 (†)
0 0 2 2 2

0 0 0−0
T =0+ x + 1x2 = 0 · (1 + x) + 0 · (1 − x) + 1 · x2 X2
0 1 2

0 1 1−1
T =0+ x + 0x2 = 0 · (1 + x) + 0 · (1 − x) + 0 · x2 X2
1 0 2

0 1 1+1 1 1
T =0+ x + 0x2 = · (1 + x) − · (1 − x) + 0 · x2 X2 (‡)
−1 0 2 2 2
representation: 1
0 0 21

2
 1 0 0 − 1  .X 4
2 2
0 1 0 0
The coefficients in (†) are found as follows. Let u, v, w ∈ R such that
u(1 + x) + v(1 − x) + wx2 = 1
i.e. u + v = 1, u − v = 0 and w = 0. These linear equations are easily solved to yield
u = 1/2, v = 1/2 and w = 0.
The coefficients in (‡) are found as follows. Let u, v, w ∈ R such that

u(1 + x) + v(1 − x) + wx2 = x
i.e. u + v = 0, u − v = 1 and w = 0. These linear equations are easily solved to yield
u = 1/2, v = −1/2 and w = 0.
223
(4.3) Determine the range R(T ) of T . Is T onto? In other words, is it true that R(T ) = P2 ? (4)

a b
R(T ) = T : a, b, c, d ∈ R X2
c d

b−c 2
= a+ x + dx : a, b, c, d ∈ R
2
= a + b0 x + dx2 : a, b0 , d ∈ R (Setting b0 = (b − c)/2)

= P2 .
Thus T is onto.X2

a b a b
ker(T ) = : a, b, c, d ∈ R, T = 0 + 0x + 0x X2
2
c d c d

a b b−c 2 2
= : a, b, c, d ∈ R, a + x + dx = 0 + 0x + 0x
c d 2

a b
= : a, b, c, d ∈ R, a = 0, b = c, d = 0
c d

0 b
= : b ∈ R,
b 0

0 1
= b : b ∈ R, .
1 0
Thus we have a one-dimensional space and the nullity of T is 1. X2
224
MAT2611/101/3/2017

Question paper
(1.1) Find (2)

1 1 1 0
S := span , .
1 1 0 −1

1 0
(1.2) Is ∈ S? Explain. (3)
0 1
(1.3) Let (8)

1 0
X := A ∈ M22 : A = ⊂ M22 .
1 0
Show that X is a vector subspace of M22 .
(1.4) What is the dimension of the vector subspace X? (4)
225

   
x1 y1
x2  y2  4
hx, yi := x1 y1 + x2 y2 + x3 y3 + 3x4 y4 , x3  , y = y3  ∈ R
x=   
x4 y4

     
1 −1 −1
0 0 1
 ,  ,  
0 0 0
1 1 1

     

 1 −1 −1 
     
0 ,  0  ,  1 

 0  0   0  
 
1 1 1
 

2 −1 −1
A = −1 2 −1 .
−1 −1 2
λ(λ − 3)2 = 0.
(3.4) Is the matrix (5)

1 1
0 1
diagonalizable? Motivate your answer.
226
MAT2611/101/3/2017
T (a + bx + cx2 ) = c + ax + bx2
(4.2) Find the matrix representation of T relative to the basis (11)
{ 1 + x + x2 , 1 − x, 1 + x − 2x2 }
ordered from left to right.
(4.3) The transform T has the eigenvalue 1. Find the corresponding eigenspace. (5)
227
Solution
Please note: any fundamental error is grounds for no marks being awarded for an answer.
(1.1) Find (2)

1 1 1 0
S := span , .
1 1 0 −1
We have

1 1 1 0 a+b a
S= a +b : a, b ∈ R = : a, b ∈ R .X2
1 1 0 −1 a a−b

1 0
(1.2) Is ∈ S? Explain. (3)
0 1
Solving
1 0 a+b a
=
0 1 a a−b
b, a = 0 and 1 = a − b. Clearly b = 1 and b = −1 is impossible to satisfy.
yields 1 = a +
1 0
Hence / S.X3
∈
0 1
(1.3) Let (8)

1 0
X := A ∈ M22 : A = ⊂ M22 .
1 0
Show that X is a vector subspace of M22 .
We have
0 0 1 0 0 0
= ⇒ ∈ X.
0 0 1 0 0 0
Thus X is non-empty. X2 Let A, B ∈ X (i.e. A, B ∈ M22 and A [ 11 ] = B [ 11 ] = [ 00 ]).
1. A + B ∈ M22 (since M22 is a vector space)

1 1 1 0 0 0
(A + B) =A +B = + = ⇒ A + B ∈ X.X3
1 1 1 0 0 0
2. Let k ∈ R. kA ∈ M22 (since M22 is a vector space)

1 1 0 0
(kA) =k A =k = ⇒ kA ∈ X.X3
1 1 0 0
Thus this set forms a vector subspace.
228
MAT2611/101/3/2017
(1.4) What is the dimension of the vector subspace X? (4)

Let
a b
A= ∈ X,
c d
a, b, c, d ∈ R. It follows that

a b 1 a+b 0
= =
c d 1 c+d 0
so that a = −b, c = −d and

a −a
X := : a, c ∈ R .
c −c
There are 2 free parameters and the dimension is 2.X4
229

   
x1 y1
x2  y2  4
hx, yi := x1 y1 + x2 y2 + x3 y3 + 3x4 y4 , x3  , y = y3  ∈ R
x=   
x4 y4
     
x1 y1 z1
x2  y2  z2  4
x3  , y = y3  , z = z3  ∈ R
x=     
x4 y4 z4
1. hx, yi = x1 y1 + x2 y2 + x3 y3 + 3x4 y4 = y1 x1 + y2 x2 + y3 x3 + 3y4 x4 = hy, xi X2
2. hx + z, yi = (x1 + z1 )y1 + (x2 + z2 )y2 + (x3 + z3 )y3 + 3(x4 + z4 )y4
= x1 y1 + z1 y1 + x2 y2 + z2 y2 + x3 y3 + z3 y3 + 3x4 y4 + 3z4 y4
= x1 y1 + x2 y2 + x3 y3 + 3x4 y4 + z1 y1 + z2 y2 + z3 y3 + 3z4 y4 = hx, yi + hz, yi
X4
3. hkx, yi = (kx1 )y1 + (kx2 )y2 + (kx3 )y3 + 3(kx4 )y4 = k(x1 y1 + x2 y2 + x3 y3 + 3x4 y4 ) =
khx, yiX2
4. hx, xi = x21 + x22 + x23 + 3x24 ≥ 0X2 so that hx, xi ≥ 0 and hx, xi = 0 if and only if
x1 = x2 = x3 = x4 = 0 (since x21 , x22 , x23 , x24 ≥ 0)X2 , i.e. x = 0.

     
1 −1 −1
0 0 1
 ,  ,  
0 0 0
1 1 1
     
1 −1 −1
0 0 1 2
0 + b  0  + c  0  = 0X
a     
1 1 1
a−b−c=0
c=0
0=0
a+b+c=0
230
MAT2611/101/3/2017
Adding the fourth equation to the first provides a = 0. The second equation provides

     

 1 −1 −1 
     
0 0
 ,  ,  1 

 0  0   0  
 
1 1 1
 
Let      
1 −1 −1
0 0 1
u1 := 
0 , u2 :=  0  , u3 :=  0  .
    
1 1 1
 
1
0
0 X
v1 := u1 =  
1
hv1 , v1 i = 12 + 02 + 02 + 3 · 12 = 4X
hu2 , v1 i = −1 · 1 + 0 · 0 + 0 · 0 + 3 · 1 · 1 = 2X
     
−1 1 −3
hu2 , v1 i  0  2 0 1  0 
   
 2
v2 := u2 − v1 X = 
 0  − 4 0 = 2  0  X
hv1 , v1 i
1 1 1
1
hv2 , v2 i = ((−3)2 + 02 + 0 · 0 + 3 · 12 ) = 3X
4
hu3 , v1 i = −1 · 1 + 1 · 0 + 0 · 0 + 3 · 1 · 1 = 2X

3 1
hu3 , v2 i = −1 · − + 1 · 0 + 0 · 0 + 3 · 1 · = 3X
2 2
hu3 , v1 i hu3 , v2 i
v3 := u3 − v1 − v2 X
hv1 , v1 i hv2 , v2 i
     
−1 1 −3
 1  2 0 3 1  0 
= 0  − 4 0 − 3 · 2  0 
    
1 1 1
 
0
1 2
=0 .X

231
     

 1 −3 0 
  1  
 ,  0 ,
0 1  2
  .X
 0 2  0  0 
 
1 1 0
 

2 −1 −1
A = −1 2 −1 .
−1 −1 2
λ(λ − 3)2 = 0.
    
1 0 0 2 −1 −1 λ−2 1 1
det λ 0 1 0 − −1 2 −1 X2 = 1 λ−2 1
0 0 1 −1 −1 2 1 1 λ−2
= (λ − 2)3 + 1 + 1 − (λ − 2) − (λ − 2) − (λ − 2)
= λ3 − 6λ2 + 9λ = λ(λ − 3)2 = 0X4 .

    
−2 1 1 x 0
 1 −2 1  y  = 0 X2
1 1 −2 z 0
for x, y, z ∈ R. The resulting equations are −2x + y + z = 0, x − 2y + z = 0 and

x + y − 2z = 0. Obviously x = 2y − z and y = 2x − z = 4y − 3z. Thus x = y = z. We
find the 1-dimensional eigenspace
      
 x   1 
x : x ∈ R = x 1 : x ∈ R .X4
x 1
   

 1 
1 .X2
1
 
232
MAT2611/101/3/2017

    
1 1 1 x 0
1 1 1 y  = 0 X2
1 1 1 z 0
for x, y, z ∈ R. The resulting equation is x + y + z = 0, i.e. x = −y − z. Thus we find

the 2-dimensional eigenspace
        
 −y − z   −1 −1 
 y  : y, z ∈ R = y  1  + z  0  : x, y ∈ R .X4
z 0 1
   

 −1 −1 
 1  ,  0  .X2
0 1
 

(i.e. 1 for λ = 0 and 2 for λ = 3).
Alternative:
Yes, since A is symmetric.
(3.4) Is the matrix (5)

1 1
0 1
diagonalizable? Motivate your answer.
Since the matrix is upper triangular, the eigenvalues are the diagonal entries, i.e. 1
(twice)X2 . Solving the eigenvalue equation

1 1 x x
=
0 1 y y
for x, y ∈ R yields x + y = x (and y = y), i.e. y = 0. Thus the 1-dimensional eigenspace

is given by
x
: x ∈ R X2
0
and the algebraic multiplicity (i.e. 2) is not equal to the geometric multiplicity (i.e 1).
Thus the matrix is not diagonalizable.X
233
T (a + bx + cx2 ) = c + ax + bx2

Let k, a, b, c, α, β, γ ∈ R. Using the definition of P2 and T we find
•
T k(a + bx + cx2 ) = T (ka) + (kb)x + (kc)x2

= (kc) + (ka)x + (kb)x2

= k(c + ax + bx2 ) = kT a + bx + cx2 .X2

•
T (a + bx + cx2 ) + (α + βx + γx2 ) = T (a + α) + (b + β)x + (c + γ)x2

= (c + γ) + (a + α)x + (b + β)x2
= (c + ax + bx2 ) + (γ + αx + βx2 )
= T a + bx + cx2 + T α + βx + γx2 .X2

(4.2) Find the matrix representation of T relative to the basis (11)

{ 1 + x + x2 , 1 − x, 1 + x − 2x2 }
ordered from left to right.
From
T (1 + x + x2 ) = 1 + x + x2
= 1 · (1 + x + x2 ) + 0 · (1 − x) + 0 · (1 + x − 2x2 )X3
T (1 − x) = T (1 − x + 0 · x2 ) = x − x2

1 1
2
= 0 · (1 + x + x ) + − · (1 − x) + · (1 + x − 2x2 )X3 (†)
2 2
T (1 + x − 2x2 ) = −2 + x + x2

3 1
2
= 0 · (1 + x + x ) + − · (1 − x) + − · (1 + x − 2x2 )X3 (‡)
2 2
1 0 0 
 
 1 3  2
.X
0 − − 
2 2


 1 1
0 −
2 2
234
MAT2611/101/3/2017
The coefficients in (†) are found as follows. Let u, v, w ∈ R such that
u(1 + x + x2 ) + v(1 − x) + w(1 + x − 2x2 ) = x − x2
i.e. u + v + w = 0, u − v + w = 1 and u − 2w = −1. These linear equations are easily

solved to yield u = 0, v = −1/2 and w = 1/2.
The coefficients in (‡) are found as follows. Let u, v, w ∈ R such that
u(1 + x + x2 ) + v(1 − x) + w(1 + x − 2x2 ) = −2 + x + x2
i.e. u + v + w = −2, u − v + w = 1 and u − 2w = 1. These linear equations are easily

solved to yield u = 0, v = −3/2 and w = −1/2.
The corresponding eigenvalue equation is
T (a + bx + cx2 ) = c + ax + bx2 = a + bx + cx2 X2
where a, b, c ∈ R from which follows a = b = cX. The eigenspace is
a + ax + ax2 : a ∈ R .X2

235
Question paper
Let n ∈ N. Consider the set of n × n symmetric matrices over R with the usual addition and
multiplication by a scalar.
(1.1) Show that this set with the given operations is a vector subspace of Mnn . (6)
(1.2) What is the dimension of this vector subspace? (4)
(1.3) Find a basis for the vector space of 2 × 2 symmetric matrices. (6)

   
x1 y1
  
x3 y3
(2.2) Are the vectors       (10)

1 1 1
1 ,  1  , −1
1 −1 −1

     
 1 1 1 
1 ,  1  , −1
1 −1 −1
 
to find an orthogonal basis with respect to the inner product defined in 2.1 for the span
of this subset.

1 0 2
A = 0 1 0 .
2 0 1
2
(λ − 1)(λ − 3) = 0.
236
MAT2611/101/3/2017
(3.2) Find an orthogonal matrix P which diagonalizes A. (16)
(3.3) Find An (for n ∈ N) as a matrix. (6)
(4.1) Is span{ 1 + x, x + x2 , x2 + x3 , x3 + 1 } = P3 ? Motivate your answer. (4)
(4.2) Let D : P3 → P3 be the differentiation operator
D(a0 + a1 x + a2 x2 + a3 x3 ) = a1 + 2a2 x + 3a3 x2 .
(a) Find the matrix representation of D relative to the basis { 1, x, x2 , x3 } using the (8)
coefficient ordering
 
a0
a1 
a0 + a1 x + a2 x2 + a3 x3 → 
a2  .

a3
(b) Find the kernel and range of D. (8)
237
Solution
Let n ∈ N. Consider the set of n × n symmetric matrices over R with the usual addition and
multiplication by a scalar.
(1.1) Show that this set with the given operations is a vector subspace of Mnn . (6)
Obviously the set of n × n symmetric matrices is a subset of Mnn . X2
Let A and B be n × n symmetric matrices over R, i.e. A = AT and B = B T . Obviously
the n × n zero matrix is symmetric, i.e. the set is not empty. The properties of the
transpose provide
1. (A + B)T = AT + B T = A + B i.e. A + B is symmetric. X2
2. Let k ∈ R. Then (kA)T = k(AT ) = kA which is also symmetric. X2
(1.2) What is the dimension of this vector subspace? (4)
Let  
a1,1 a1,2 a1,3 . . . a1,n
 a1,2 a2,2 a2,3 . . . a2,n 
 
A =  a1,3 a2,3 a3,3 . . . a3,n 


 .. .. ... .. 
 . . . 
a1,n a2,n a3,n . . . an,n
be an element of this vector space, with the free parameters a1,1 , . . . , a1,n ∈ R, a2,2 , . . . , a2,n ∈
R, a3,3 , . . . a3,n ∈ R,. . . . . . , an,n ∈ R. We sum the number of free parameters in each
row to obtain
n
n(n + 1)
X4
X
n + (n − 1) + (n − 2) + . . . + 1 = j=
j=1
2
for the dimension.
(1.3) Find a basis for the vector space of 2 × 2 symmetric matrices. (6)
A typical 2 × 2 symmetric matrix over R has the form

a b
b c
where a, b, c ∈ R. Thus we have the vector space

a b 1 0 0 1 0 0
: a, b, c ∈ R = a +b +c : a, b, c ∈ R .
b c 0 0 1 0 0 1
The dimension of the vector space is 3. An obvious choice for a basis is

1 0 2 0 0 2 0 1 2
0 0
X, 0 1 X, 1 0 X .
Other choices include

1 0 1 0 0 1 1 1 1 0 1 −1
, , and , , .
0 1 0 −1 1 0 1 1 0 −1 −1 1
238
MAT2611/101/3/2017

   
x1 y1
hx, yi := x1 y1 + 2x2 y2 + x3 y3 , x = x2  , y = y2  ∈ R3
x3 y3
     
x1 y1 z1
x = x2  , y = y2  , z = z2  ∈ R3
x3 y3 z3
1. hx, yi = x1 y1 + 2x2 y2 + x3 y3 = y1 x1 + 2y2 x2 + y3 x3 = hy, xi X2
2. hx + z, yi = (x1 + z1 )y1 + 2(x2 + z2 )y2 + (x3 + z3 )y3

= x1 y1 + z1 y1 + 2x2 y2 + 2z2 y2 + x3 y3 + z3 y3
= x1 y1 + 2x2 y2 + x3 y3 + z1 y1 + 2z2 y2 + z3 y3 = hx, yi + hz, yi X4
3. hkx, yi = (kx1 )y1 + 2(kx2 )y2 + (kx3 )y3 = k(x1 y1 + 2x2 y2 + x3 y3 ) = khx, yiX2
4. hx, xi = x21 + 2x22 + x23 ≥ 0X2 so that hx, xi ≥ 0 and hx, xi = 0 if and only if
x1 = x2 = x3 = 0 (since x21 , x22 , x23 ≥ 0)X2 , i.e. x = 0.
(2.2) Are the vectors       (10)

1 1 1
1 ,  1  , −1
1 −1 −1
     
1 1 1
a 1 + b 1 + c −1 = 0X2
    
1 −1 −1
where a, b, c ∈ R. Thus we have the equation
a+b+c = 0
a+b−c = 0
a−b−c = 0
Adding the first and third equation provides a = 0X2 . Adding the first and sec-
ond equation provides b = −a = 0X2 . Adding the second and third equation provides
239
c = a = 0X2 . This is the only solution. Thus these vectors are linearly independent.X2
Alternative:
The coefficient matrix is  
1 1 1
1 1 −1
1 −1 −1
which has a determinant of -2, i.e. the determinant is non-zero from which follows that
these vectors are linearly independent.

     
 1 1 1 
1 ,  1  , −1
1 −1 −1
 
to find an orthogonal basis with respect to the inner product defined in 2.1 for the span
of this subset.
Let      
1 1 1
u1 := 1 , u2 := 1 , u3 := −1 .
    
1 −1 −1
 
1
v1 := u1 = 1
1
hv1 , v1 i = 12 + 2 · 12 + 12 = 4X2
     
1 1 1
hu2 , v1 i 2  1  4
v2 := u2 − v1 = 1 −
  1 = 1 X
hv1 , v1 i 4 2
−1 1 −3
2 2 2
1 1 3
hv2 , v2 i = +2 + − = 3X2
2 2 2
hu3 , v1 i hu3 , v2 i
v3 := u3 − v1 − v2
hv1 , v1 i hv2 , v2 i
     
1 1 1
−2   2 1 1   2
= −1 − 1 X − · 1 X
4 3 2
−1 1 −3
 
4
1  2
= −2 X .
3
0
     
 1 1
1
1
4 
1 ,  1  , −2 .
2 3
1 −3 0
 
240
MAT2611/101/3/2017

1 0 2
A = 0 1 0 .
2 0 1
(λ2 − 1)(λ − 3) = 0.
    
1 0 0 1 0 2 λ−1 0 −2
det λ 0 1 0 − 0 1 0 X2 = 0 λ−1 0 = (λ − 1)3 − 4(λ − 1)X2
0 0 1 2 0 1 −2 0 λ−1
= (λ − 1)((λ − 1)2 − 4) = (λ − 1)(λ − 3)(λ + 1) = 0X2 .
Cofactor expansion along any row or column could also be used.
(3.2) Find an orthogonal matrix P which diagonalizes A. (16)

From the characteristic equation we obtain the eigenvalues -1X2 , 1X2 and 3X2 . For
the eigenspace corresponding to the eigenvalue 1 we solve
    
1 0 2 x x
0 1 0 y  = y 
2 0 1 z z
for x, y, z ∈ R. The resulting equations are x + 2z = x, 2x + z = z and y = y which is

satisfied identically. Thus we find the 1-dimensional eigenspace
  
 0 
y  X2 : y ∈ R .
0
 
For the eigenspace corresponding to the eigenvalue -1 we solve

    
1 0 2 x x
0 1 0 y  = − y 
2 0 1 z z
for x, y, z ∈ R. The resulting equations are x + 2z = −x, 2x + z = −z and y = −y i.e.

z = −x and y = 0. Thus we find the 1-dimensional eigenspace
  
 x 
 0  X2 : x ∈ R .
−x
 
241
    
1 0 2 x x
0 1 0 y  = 3 y 
2 0 1 z z
for x, y, z ∈ R. The resulting equations are x + 2z = 3x, 2x + z = 3z and y = 3y i.e.

z = x and y = 0. Thus we find the 1-dimensional eigenspace
  
 x 
 0  X2 : x ∈ R .
x
 
Choosing orthonormal representative eigenvectors

     
0 1 1
1 , √1  0  , √1 0
0 2 −1 2 1
from each eigenspace we find that
0 √12 √1
 
2
P = 1 0 0  X4
0 − √12 √1
2
diagonalizes A, i.e.  
1 0 0
P T AP = 0 −1 0 .
0 0 3
Of course, other P could also be used (for example by rearranging columns or by mul-
tiplying a column by -1).
(3.3) Find An (for n ∈ N) as a matrix. (6)

From  
1 0 0
(P T AP )n = 0 (−1)n 0  X2
0 0 3n
we find
 n
3 + (−1)n 3n − (−1)n


1 0 0
 0
n n T
 2 2 
 4
A = P 0 (−1) 0 P = 0 1 0  .X
  
 3n − (−1)n 3n + (−1)n 
0 0 3n
0
2 2
242
MAT2611/101/3/2017
(4.1) Is span{ 1 + x, x + x2 , x2 + x3 , x3 + 1 } = P3 ? Motivate your answer. (4)

No, since
a0 + a1 x + a2 x2 + a3 x3 = b0 (1 + x) + b1 (x + x2 ) + b2 (x2 + x3 ) + b3 (x3 + 1)
gives b0 + b1 = a1 , b1 + b2 = a2 , b2 + b3 = a3 and b3 + b0 = a0 so that
b 0 = a1 − b 1 = a1 − a2 + b 2 = a1 − a2 + a3 − b 3 = a1 − a2 + a3 − a0 + b 0 ⇒ a1 + a3 = a0 + a2 .
/ span{ 1 + x, x + x2 , x2 + x3 x3 + 1 }. X4
For example x ∈
Alternative:
The augmented coefficient matrix for the above equation is
   
1 0 0 1 : a0 ∼ 1 0 0 1 : a0
1 1
 0 0 : a1  R2 − R1 0 1
  0 −1 : a1 − a0 

0 1 1 0 : a2  ∼ 0 1 1 0 : a2 
0 0 1 1 : a3 ∼ 0 0 1 1 : a3
 
∼ 1 0 0 1 : a0
∼ 0 1
 0 −1 : a1 − a0  
R3 − R2 0 0 1 1 : a2 − a1 + a0 
∼ 0 0 1 1 : a3
 
∼ 1 0 0 1 : a0
∼ 0 1
 0 −1 : a1 − a0 

∼ 0 0 1 1 : a2 − a1 + a0 
R4 − R3 0 0 0 0 : a3 − a2 + a1 − a0
Clearly the last equation cannot be satisfied if a3 − a2 + a1 − a0 6= 0.
Alternative:
Noting that the determinant of the coefficient matrix is zero, the dimension of span{ 1 +
x, x + x2 , x2 + x3 , x3 + 1 } is less than 4, while the dimension of P3 is 4 : i.e. the two
vector spaces cannot be the same.
243
(4.2) Let D : P3 → P3 be the differentiation operator
D(a0 + a1 x + a2 x2 + a3 x3 ) = a1 + 2a2 x + 3a3 x2 .
(a) Find the matrix representation of D relative to the basis { 1, x, x2 , x3 } using the (8)
coefficient ordering
 
a0
a1 
a0 + a1 x + a2 x2 + a3 x3 → 
a2  .

a3
From
D(1) = D(1 · 1 + 0 · x + 0 · x2 + 0 · x3 ) = 0 = 0 · 1 + 0 · x + 0 · x2 + 0 · x3
D(x) = D(0 · 1 + 1 · x + 0 · x2 + 0 · x3 ) = 1 = 1 · 1 + 0 · x + 0 · x2 + 0 · x3
D(x2 ) = D(0 · 1 + 0 · x + 1 · x2 + 0 · x3 ) = 2x = 0 · 1 + 2 · x + 0 · x2 + 0 · x3
D(x3 ) = D(0 · 1 + 0 · x + 0 · x2 + 1 · x3 ) = 3x2 = 0 · 1 + 0 · x + 3 · x2 + 0 · x3
the coefficients of the basis elements in each equation provide the columns of the
matrix representation:  
0 1 0 0 X2
2
 X2 .
0 0 2 0
3 X2

0 0 0
0 0 0 0 X
(b) Find the kernel and range of D. (8)

Solving
D(a0 + a1 x + a2 x2 + a3 x3 ) = a1 + 2a2 x + 3a3 x2 = 0X2
we find a1 = a2 = a3 = 0 so that
ker(D) = {a0 · 1 + 0 · x + 0 · x2 + 0 · x3 : a0 ∈ R} ≡ {a0 : a0 ∈ R} ≡ R.X2
We have
D a0 + a1 x + a2 x2 + a3 x3 = a1 + 2a2 x + 3a3 x2

where a0 , a1 , a2 , a3 ∈ R. Thus
R(D) = {a1 + 2a2 x + 3a3 x2 + 0 · x3 : a1 , a2 , a3 ∈ R}

≡ {a01 + a02 x + a03 x2 : a01 , a02 , a03 ∈ R} ≡ P2 X4
where a01 = a1 , a02 = 2a2 and a03 = 3a3 .
244
MAT2611/101/3/2017

Question paper
(1.1) Is span{ 1, 1 + x, x + x2 , x2 + x3 , x3 + 1 } = P3 ? Motivate your answer. (7)
(1.2) Let a ∈ R and (8)

Za := { p(x) ∈ P3 : p(a) = 0 } ⊂ P3 .
Show that Za is a vector subspace of P3 .
(1.3) What is the dimension of the vector subspace Za ? (3)

   
x1 y1
hx, yi := x1 y1 + x2 y2 + 3x3 y3 , x = x2 , y = y2  ∈ R3
  
x3 y3
(2.2) Are the vectors       (10)

1 −1 −1
0 ,  0  ,  1 
1 1 1

     
 1 −1 −1 
0 ,  0  ,  1 
1 1 1
 
245

1 0 1
A = 0 2 0 .
1 0 1
λ(λ − 2)2 = 0.

0 1
(3.4) Is diagonalizable? Motivate your answer. (5)
0 0
Let T : M22 → M22 be the transpose operation

a b a c
T =
c d b d
(4.2) Find the matrix representation of T relative to the standard basis (8)

1 0 0 1 0 0 0 0
, , ,
0 0 0 0 1 0 0 1
using the coefficient ordering
 
a
a b 1 0 0 1 0 0 0 0  b
=a +b +c +d →
c .

c d 0 0 0 0 1 0 0 1
d
(4.3) Find the kernel of T . (4)
(4.4) Is T invertible? (2)

Explain using only your answer for 4.3.
246
MAT2611/101/3/2017
Solution
(1.1) Is span{ 1, 1 + x, x + x2 , x2 + x3 , x3 + 1 } = P3 ? Motivate your answer. (7)

Yes, since
a0 + a1 x + a2 x2 + a3 x3 = b0 + b1 (1 + x) + b2 (x + x2 ) + b3 (x2 + x3 ) + b4 (x3 + 1)X2
gives b0 + b1 + b4 = a0 , b1 + b2 = a1 , b2 + b3 = a2 , and b3 + b4 = a3 X2 so that
b 3 = a3 − b 4 , b2 = a2 − b3 = a2 − a3 + b4 , b1 = a1 − b2 = a1 − a2 + a3 − b4 ,
b0 = a0 − b1 − b4 = a0 − a1 + a2 − a3 .X2
where b4 is a free parameter. Thus we found a solution.X
Alternative:
The augmented coefficient matrix for the above equation is
   
1 1 0 0 1 : a0 ∼ 1 1 0 0 1 : a0
0 1 1 0 0 : a1  ∼ 0 1 1 0 0 : a1 
   
0 0 1 1 0 : a2 R3 − R4 0 0 1 0 −1 : a2 − a3 
0 0 0 1 1 : a3 ∼ 0 0 0 1 1 : a3
 
∼ 1 1 0 0 1 : a0
R2 − R3  1 0 0 1
0 : a1 − a2 + a3 

∼ 0 0 1 0 −1 : a2 − a3 
∼ 0 0 0 1 1 : a3
 
R1 − R2 1 0 0 0 0 : a0 − a1 + a2 − a3
∼ 0 1 0 0 1
 : a1 − a2 + a3  
∼ 0 0 1 0 −1 : a2 − a3 
∼ 0 0 0 1 1 : a3
which yields the same solution as above.
(1.2) Let a ∈ R and (8)
Za := { p(x) ∈ P3 : p(a) = 0 } ⊂ P3 .
Show that Za is a vector subspace of P3 .
We have p1 (x) := −a + x ∈ Za since p1 (x) = −a + x = −a · 1 + 1 · x + 0 · x2 + 0 · x3 ∈ P3
and p1 (a) = −a + a = 0. Thus Za is non-empty (we could also have used p1 (x) :=
0+0·x+0·x2 +0·x3 ). X2 Let p(x), q(x) ∈ Za (i.e. p(x), q(x) ∈ P3 and p(a) = q(a) = 0).
1. p(x) + q(x) ∈ P3 X(since P3 is a vector space)
(p + q)(a) = p(a) + q(a) = 0 + 0 = 0X ⇒ p(x) + q(x) ∈ Za X
2. Let k ∈ R. kp(x) ∈ P3 X(since P3 is a vector space)
(kp)(a) = k p(a) = k · 0 = 0X ⇒ kp(x) ∈ Za X
247
(1.3) What is the dimension of the vector subspace Za ? (3)
Let p(x) = p0 + p1 x + p2 x2 + p3 x3 ∈ Za , p0 , p1 , p2 , p3 ∈ R. It follows that
p(a) = p0 + p1 a + p2 a2 + p3 a3 = 0 ⇒ p0 = −p1 a − p2 a2 − p3 a3 X2
so that
Za := { (−p1 a − p2 a2 − p3 a3 ) + p1 x + p2 x2 + p3 x3 : p1 , p2 , p3 ∈ R }.
There are 3 free parameters and the dimension is 3.X

   
x1 y1
hx, yi := x1 y1 + x2 y2 + 3x3 y3 , x = x2 , y = y2  ∈ R3
  
x3 y3
     
x1 y1 z1
3
x= x 2 ,y = y2 , z = z2 ∈ R
     
x3 y3 z3
1. hx, yi = x1 y1 + x2 y2 + 3x3 y3 = y1 x1 + y2 x2 + 3y3 x3 = hy, xi X2
2. hx + z, yi = (x1 + z1 )y1 + (x2 + z2 )y2 + 3(x3 + z3 )y3

= x1 y1 + z1 y1 + x2 y2 + z2 y2 + 3x3 y3 + 3z3 y3
= x1 y1 + x2 y2 + 3x3 y3 + z1 y1 + z2 y2 + 3z3 y3 = hx, yi + hz, yi X4
3. hkx, yi = (kx1 )y1 + (kx2 )y2 + 3(kx3 )y3 = k(x1 y1 + x2 y2 + 3x3 y3 ) = khx, yiX2
4. hx, xi = x21 + x22 + 3x23 ≥ 0X2 so that hx, xi ≥ 0 and hx, xi = 0 if and only if
x1 = x2 = x3 = 0 (since x21 , x22 , x23 ≥ 0)X2 , i.e. x = 0.
248
MAT2611/101/3/2017
(2.2) Are the vectors       (10)

1 −1 −1
0 ,  0  ,  1 
1 1 1
     
1 −1 −1
a 0 + b 0 + c 1  = 0X2
    
1 1 1
where a, b, c ∈ R. Thus we have the equation
a−b−c=0
c=0
a+b+c=0
Adding the third equation to the first provides a = 0X2 . The second equation provides
c = 0.X2 Inserting the solutions for a and c into the first equation yields b = 0.X2 This
is the only solution. Thus these vectors are linearly independent.X2
Alternative:
The coefficient matrix is  
1 −1 −1
0 0 1
1 1 1
which has a determinant of -2, i.e. the determinant is non-zero from which follows that
these vectors are linearly independent.

     
 1 −1 −1 
0 ,  0  ,  1 
1 1 1
 
Let      
1 −1 −1
u1 := 0 , u2 := 0 , u3 := 1  .
    
1 1 1
249
 
1
v1 := u1 = 0 X
1
hv1 , v1 i = 12 + 02 + 3 · 12 = 4X
hu2 , v1 i = −1 · 1 + 0 · 0 + 3 · 1 · 1 = 2X
     
−1 1 −3
hu2 , v1 i 2  1  2
v2 := u2 − v1 X = 0 −
  0 = 0 X
hv1 , v1 i 4 2
1 1 1
1
hv2 , v2 i = ((−3)2 + 02 + 3 · 12 ) = 3X
4
hu3 , v1 i = −1 · 1 + 1 · 0 + 3 · 1 · 1 = 2X

3 1
hu3 , v2 i = −1 · − + 1 · 0 + 3 · 1 · = 3X
2 2
hu3 , v1 i hu3 , v2 i
v3 := u3 − v1 − v2 X
hv1 , v1 i hv2 , v2 i
     
−1 1 −3
2  3 1 
= 1 −
  0 − · 0
4 3 2
1 1 1
 
0
= 1 .X2
0
     
 1 −3 0 
0 , 1  0  , 1 .X2
2
1 1 0
 

1 0 1
A = 0 2 0 .
1 0 1
λ(λ − 2)2 = 0.
    
1 0 0 1 0 1 λ−1 0 −1
det λ 0 1 0 − 0 2 0
      X = 0 λ − 2 0 = (λ − 1)2 (λ − 2) − (λ − 2)X2
2
0 0 1 1 0 1 −1 0 λ−1
= (λ − 2)((λ − 1)2 − 1) = (λ − 2)2 λ = 0X2 .
250
MAT2611/101/3/2017

    
1 0 1 x 0
0 2 0 y  = 0
1 0 1 z 0
for x, y, z ∈ R. The resulting equations are x + z = 0, 2y = 0 and x + z = 0.X2

Obviously y = 0 and z = −x. Thus we find the 1-dimensional eigenspace
      
 x   1 
 0  : x ∈ R = x 0  : x ∈ R .
−x −1
   

 1 
 0  .X 2
−1
 

    
1 0 1 x 2x
0 2 0 y  = 2y 
1 0 1 z 2z
for x, y, z ∈ R. The resulting equations are x + z = 2x, 2y = 2y and x + z = 2z X2 , i.e.

x = z and y = y. Thus we find the 2-dimensional eigenspace
        
 x   1 0 
y  : x, y ∈ R = x 0 + y 1 : x, y ∈ R .
x 1 0
   

 1 0 
0 , 1 .X2
1 0
 

(i.e. 1 for λ = 0 and 2 for λ = 2).
Alternative:
Yes, since A is symmetric.
251

0 1
(3.4) Is diagonalizable? Motivate your answer. (5)
0 0
Since the matrix is upper triangular, the eigenvalues are the diagonal entries, i.e. 0
(twice)X2 . Solving the eigenvalue equation

0 1 x 0
=
0 0 y 0
for x, y ∈ R yields y = 0. Thus the 1-dimensional eigenspace is given by

x
: x ∈ R X2
0
and the algebraic multiplicity (i.e. 2) is not equal to the geometric multiplicity (i.e 1).
Thus the matrix is not diagonalizable.X
Let T : M22 → M22 be the transpose operation

a b a c
T =
c d b d

Let k, a, b, c, d, α, β, γ, δ ∈ R. Using the definition of M22 and T we find
•

a b ka kb ka kc a c a b
T k =T = =k = kT .X 2
c d kc kd kb kd b d c d
•

a b α β a+α b+β a+α c+γ
T + =T =
c d γ δ c+γ d+δ b+β d+δ

a c α γ a b α β
= + =T +T .X2
b d β δ c d γ δ
252
MAT2611/101/3/2017
(4.2) Find the matrix representation of T relative to the standard basis (8)

1 0 0 1 0 0 0 0
, , ,
0 0 0 0 1 0 0 1
using the coefficient ordering

 
a
a b 1 0 0 1 0 0 0 0 b
=a +b +c +d →
c .

c d 0 0 0 0 1 0 0 1
d
From

1 0 1 0 1 0 0 1 0 0 0 0
T = =1· +0· +0· +0·
0 0 0 0 0 0 0 0 1 0 0 1

0 1 0 0 1 0 0 1 0 0 0 0
T = =0· +0· +1· +0·
0 0 1 0 0 0 0 0 1 0 0 1

0 0 0 1 1 0 0 1 0 0 0 0
T = =0· +1· +0· +0·
1 0 0 0 0 0 0 0 1 0 0 1

0 0 0 0 1 0 0 1 0 0 0 0
T = =0· +0· +0· +1·
0 1 0 1 0 0 0 0 1 0 0 1
1 0 0 0 X2
0 0 1 0 X2
0 1 0 0 X2 .
 
2
0 0 0 1 X
(4.3) Find the kernel of T . (4)

Solving
a b a c 0 0
T
c d
=
b d
=
0 0
X2
we find a = b = c = d = 0 so that

0 0
ker(T ) = .X 2
0 0
(4.4) Is T invertible? (2)

Explain using only your answer for 4.3.
Yes, because T : M22 → M22 and

0 0
ker(T ) =
0 0
X2
consists only of the zero vector in M22 .
253
The corresponding eigenvalue equation is

a b a c a b
T
c d
=
b d
=
c d
X2
where a, b, c, d ∈ R from which follows b = cX. The eigenspace is

a b
: a, b, d ∈ R X2
b d
(the 2 × 2 symmetric matrices.)
254
MAT2611/101/3/2017

Question paper
255
256
MAT2611/101/3/2017
257
Solution
Q1 From the characteristic polynomial p(λ) = λ3 − λ = λ(λ − 1)(λ + 1) we find that A has the
eigenvalues 0, 1 and −1.
(a) The order of A is 3 (degree of p(λ)).

(b) No, since 0 is an eigenvalue of A.
(c) Yes, since the eigenvalues are all different.
(d) The eigenvalues of A2 are 02 = 0, 12 = 1 and (−1)2 = 1.
Q2 (a) The characteristic polynomial of A in λ is
λ + 3 −1 0
det(λI − A) = 6 λ − 2 0 = (λ + 3)(λ − 2)λ + 6λ = λ2 (λ + 1)
3 −1 λ
so that the eigenvalues of A are 0 (twice) and -1.

(b) First we consider the eigenvalue 0. Thus we solve
      
x 3 −1 0 x 0
0I − A y = 6
   −2 0   y = 0
 
z 3 −1 0 z 0
for x, y, z ∈ R. Row reduction of the augmented matrix yields

   
3 −1 0 : 0 ∼ 3 −1 0 : 0
6 −2 0 : 0 R2 − 2R1 0 0 0 : 0
3 −1 0 : 0 R3 − R1 0 0 0 : 0
so that 3x = y. Thus the eigenspace is

        
 x   1 0 
3x : x, z ∈ R = x 3 + z 0 : x, z ∈ R
z 0 1
   

   
 1 0 
3 , 0 .
0 1
 
Don’t forget to verify that these vectors are linearly independent. If they are not, use row
reduction of the matrix with these vectors as rows and the non-zero rows of the reduced
matrix will form a basis (i.e. rewrite the non-zero rows as column vectors).
(c) We found 3 linearly independent eigenvectors of A.
(d) We construct P directly from the eigenvectors we found:
 
1 0 1
3 0 2
0 1 1
258
MAT2611/101/3/2017
then the eigenvalues in D appear in the same order as the corresponding eigenvectors in P
 
0 0 0
0 0 0 .
0 0 −1
Now check that P −1 AP = D (exercise).

(e) From P −1 AP = D we find
   
099 0 0 0 0 0
P −1 A99 P = D99 =  0 099 0  = 0 0 0  = D
0 0 (−1)99 0 0 −1
so that P −1 A99 P = D and A99 = P DP −1 = A.
Q3 (a) Let x = (a, b), y = (α, β) ∈ R2 with k, a, b, α, β ∈ R. Now
T (kx) = T (ka, kb) = (ka) + (kb) = k(a + b) = kT (a, b) = kT (x)

T (x + y) = T (a + α, b + β) = (a + α) + (b + β) = (a + b) + (α + β)
= T (a, b) + T (α, β) = T (x) + T (y)
(b) Let x = (a, b) ∈ R2 with k, a, b ∈ R. Now
T (kx) = T (ka, kb) = (ka)(kb) = k 2 (ab) = k 2 T (a, b) = k 2 T (x).
Now consider a = b = 1 and k = 2, then T (kx) = 4 and kT (x) = 2. This provides a

counter example for T (kx) = kT (x). Thus T is not linear.
Q4 (a) Applying T to the basis B we find
T (1) = 1 = 1 = 1 · 1 + 0 · x + 0 · x2
x→x−1
T (x) = x = x − 1 = −1 · 1 + 1 · x + 0 · x2
x→x−1
T (x2 ) = x2 = (x − 1)2 = x2 − 2x + 1 = 1 · 1 − 2 · x + 1 · x2
x→x−1
and writing the coefficients of the basis vectors for each equation as the column vectors in
the matrix representation we find
 
1 −1 1
[T ]B = 0 1 −2 .
0 0 1
(b) We have
T (q(x)) = c0 +c1 x+c2 x2 = c0 +c1 (x−1)+c2 (x−1)2 = (c0 −c1 +c2 )+(c1 −2c2 )x+c2 x2
x→x−1
259
so that  
c0 − c1 + c2
[T (q(x))]B =  c1 − 2c2  .
c2
We also have  
c0
[q(x)]B = c1 

c2
so that     
1 −1 1 c0 c0 − c1 + c2
[T ]B [q(x)]B = 0 1 −2 c1  =  c1 − 2c2  = [T (q(x))]B .
0 0 1 c2 c2
(c) We must express each element of B 0 in terms of the elements of B, i.e. set b(x) = α · 1 +
β · x + γ · x2 and solve for α, β, γ ∈ R for each b(x) ∈ B 0 . In this case the calculation is
straightforward:
1 + x + x 2 = 1 · 1 + 1 · x + 1 · x2
2x + x2 = 0 · 1 + 2 · x + 1 · x2
x + x 2 = 0 · 1 + 1 · x + 1 · x2
Each equation’s coefficients provides the columns for the transition matrix
 
1 0 0
P = 1 2 1 .
1 1 1
(d) To calculate T in the basis B 0 we first convert B 0 to B using P , then we apply [T ]B in B,

and then convert back from B to B 0 using P −1 :
[T ]B 0 = PB 0 ←B [T ]B PB←B 0 = P −1 [T ]B P.
We apply row-reduction to the matrix P augmented with the identity to find P −1 .

   
1 0 0 : 1 0 0 ∼ 1 0 0 : 1 0 0
1 2 1 : 0 1 0R2 − R1 0 2 1 : −1 1 0
1 1 1 : 0 0 1 R3 − R1 0 1 1 : −1 0 1
 
∼ 1 0 0 : 1 0 0
R2 − R3 0 1
 0 : 0 1 −1
∼ 0 1 1 : −1 0 1
 
∼ 1 0 0 : 1 0 0
R3 − R2 0 1
 0 : 0 1 −1
∼ 0 0 1 : −1 −1 2
Thus (exercise: check that P −1 P = P P −1 = I)

 
1 0 0
P −1 =  0 1 −1 .
−1 −1 2
260
MAT2611/101/3/2017
Now
   
1 0 0 1 −1 1 1 0 0
[T ]B 0 = P −1 [T ]B P = 0
 1 −1 0 1 −2 1 2 1
−1 −1 2 0 0 1 1 1 1
    
1 −1 1 1 0 0 1 −1 0
= 0
 1 −3 1
  2 1 = −2 −1 −2 .
 
−1 0 3 1 1 1 2 3 3
Q5 (a) ∗ W must be non-empty.

∗ For all x, y ∈ W , x + y ∈ W (closure under vector addition)
∗ For all k ∈ R and x ∈ W , kx ∈ W (closure under scalar multiplication)
(b) Let 0V be the zero vector in V and 0W be the zero vector in W .
(i) The kernel of T is
ker(T ) := { v ∈ V : T (v) = 0W }.
(ii) · Since T is linear and 0x = 0W for all x ∈ W we have
T (0V ) = T (0y) = 0T (y) = 0W
for all y ∈ V . Thus 0V ∈ ker(T ), and ker(T ) is non-empty.

· Let k ∈ R and a ∈ ker(T ). Then T (ka) = kT (a) = k0W = 0W .
Thus ka ∈ ker(T ).
· Let a, b ∈ ker(T ). Then T (a + b) = T (a) + T (b) = 0W + 0W = 0W .
Thus a + b ∈ ker(T ).
(iii) T is one-to-one
if and only if
∀x, y ∈ V : T (x) = T (y) ⇔ x = y
if and only if
∀x, y ∈ V : T (x) − T (y) = 0W ⇔ x − y = 0V
if and only if
∀x, y ∈ V : T (x − y) = 0W ⇔ x − y = 0V
if and only if
∀z ∈ V : T (z) = 0W ⇔ z = 0V
if and only if
ker(T ) = { 0V }; where we used the linearity of T and substituted z := x − y.
Q6 (a) We solve       
a 1 1 1 a 0
A b = 1 0 1
     b = 0
 
c 2 −2 2 c 0
for a, b, c ∈ R. Applying row-reduction we find
     
1 1 1 : 0 R1 − R2 0 1 0 : 0 R2 1 0 1 : 0
1 0 1 : 0 ∼ 1 0 1 : 0 R1 0 1 0 : 0
2 −2 2 : 0 R3 − 2R1 0 −2 0 : 0 R3 + 2R1 0 0 0 : 0
261
so that c = −a and b = 0. Thus
      
 a   1 
nullspace(A) =  0  : a ∈ R = a  0  : a ∈ R .
−a −1
   
Since there is only one free parameter we easily find the basis
 
 1 
0
−1
 
for nullspace(A).
(b) Since rank(A) + nullity(A) = 3 (the number of columns of A) we find
rank(A) = 3 − nullity(A) = 3 − 1 = 2.
Q7 (a) We solve
w1 = av1 + bv2 , w2 = cv1 + dv2
for a, b, c, d ∈ R to find a = b = d = 1 and c = 2. We have

a c 1 2
PS←T = = .
b d 1 1
(b) We solve
v1 = αw1 + βw2 , v2 = γw1 + δw2
for α, β, γ, δ ∈ R to find α = −1, β = 1, γ = 2 and δ = −1. We have

α γ −1 2
QT ←S = = .
β δ 1 −1
−1
Notice that QT ←S = PS←T which provides another way to find QT ←S .
(c) We have
1 2 −3 7
[v]S = PS←T [v]T = = .
1 1 5 2
Q8 (a) Let k ∈ R and p = p(x), q = q(x), r = r(x) ∈ P2 .

∗ hq, pi = q(0)p(0) + q 21 p 21 + q(1)p(1)

= p(0)q(0) + p 21 q 21 + p(1)q(1) = hp, qi

∗ hkp, qi = (kp)(0)q(0) + (kp) 21 q 12 + (kp)(1)q(1)

= k p(0)q(0) + p 21 q 12 + p(1)q(1) = khp, qi

∗ hp + q, ri = (p + q)(0)r(0) + (p + q) 21 r 12 + (p + q)(1)r(1)

1 1 1 1

= p(0)r(0) + q(0)r(0) + p
2 r 2
+ q 2
r 2
+ p(1)r(1)
+ q(1)r(1)
= p(0)r(0) + p 21 r 12 + p(1)r(1) + q(0)r(0) + q 12 r 21 + q(1)r(1)

= hp, ri + hq, ri.
262
MAT2611/101/3/2017
2
∗ hp, pi = p(0)p(0) + p 21 p 21 + p(1)p(1) = p(0)2 + p 12 + p(1)2 ≥ 0

Now let p = p(x) = p0 + p1 x + p2 x2 ∈ P2 where p0 , p1 , p2 ∈ R.

2
Then hp, pi = p20 + p0 + p21 + p42 + (p0 + p1 + p2 )2 = 0
if and only if p0 = p0 + p21 + p42 = p0 + p1 + p2 = 0
if and only if p0 = 0, 2p1 + p2 = 0, p1 + p2 = 0
if and only if p0 = 0, p1 = 0, p2 = 0
where we subtracted the last equation from the middle equation.
(b) We apply the Gram-Schmidt process
v1 = (0, 1, 2)
hv1 , (0, 1, 2)i 2
v2 = (−1, 0, 1) − v1 = (−1, 0, 1) − (0, 1, 2)
hv , v i 5
1 1
2 1 1
= −1, − , = (−5, −2, 1).
5 5 5
Normalizing v1 and v2 we find the orthonormal basis

( ) ( )
v1 v2 v 1 v2 1 1
p ,p = √ ,√ = √ (0, 1, 2), √ (−5, −2, 1) .
hv1 , v1 i hv2 , v2 i 5 530 5 30
263

Tutorial Letter 101/3/2017: Linear Algebra

Uploaded by

Copyright:

Available Formats

Tutorial Letter 101/3/2017: Linear Algebra

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tutorial Letter 101/3/2017: Linear Algebra

Uploaded by

Copyright:

Available Formats

MAT2611/101/3/2017

Tutorial letter 101/3/2017

Department of Mathematical Sciences

ADDENDUM B: ASSIGNMENTS – SECOND SEMESTER .........................................................22

You must be registered on myUnisa (http://my.unisa.ac.za) to be able to submit assignments

2 PURPOSE AND OUTCOMES FOR THE MODULE

2.2.1 Vector spaces

2.2.2 Rank of a matrix

2.2.4 Eigenvalues and eigenvectors

2.2.5 Diagonalisation of matrices

2.2.6 Inner products and orthogonality

2.2.7 Gram-Schmidt algorithm

2.2.8 Orthogonal diagonalisation of symmetric matrices

2.2.9 Linear transformations

3 LECTURER(S) AND CONTACT DETAILS

Postal address: The MAT2611 Lecturers

Your prescribed textbook for this module is:

Title: Elementary Linear Algebra with Supplemental Applications

Title: Linear Algebra and Its Applications

4.2 Recommended books

4.3 Electronic reserves (e-Reserves)

4.4 Library services and resources information

The Library has compiled numerous library guides:

• find recommended reading in the print collection and e-reserves

• postgraduate information services

• how to contact the Library/find us on social media/frequently asked questions

5 STUDENT SUPPORT SERVICES

7 PRACTICAL WORK AND WORK INTEGRATED LEARNING

You must be able to do the following.

• Test a given set of vectors for linear dependence/independence.

• Find the rank and nullity of a given matrix.

You must be able to do the following.

• Calculate inner products in cases other than the dot product.

• Test vectors for orthogonality.

• Find orthogonal complements of subspaces.

• Test sets of vectors for orthogonality/orthonormality.

• Use the Gram-Schmidt process to change a basis to an orthogonal/orthonormal one.

• Find the transition matrix between two different bases.

• Find the coordinate vector of a vector with respect to a new basis.

• Decide whether or not a matrix is orthogonal.

You must be able to do the following.

• Test whether a given scalar/vector pair is an eigenvalue–eigenvector pair of a matrix.

• Find the eigenvalues and eigenvectors of a matrix.

• Determine whether or not a given matrix is diagonalisable, giving reasons.

• Find, for a diagonalisable matrix, a diagonalising matrix.

• Determine whether or not a given matrix is orthogonally diagonalisable, giving reasons.

• Find, for an orthogonally diagonalisable matrix, an orthogonal diagonalising matrix.

The student must be able to:

• Find the kernel and range of a linear transformation.

• Find the rank and nullity of a linear transformation.

• Determine whether a given linear transformation is one-to-one and/or onto.

• Find, in those cases where is is possible, the inverse of a linear transformation.

• Find the matrix of a linear transformation with respect to a given basis.

• Find the eigenvalues of a linear operator.

• Decide if a given linear transformation is an isomorphism or not, with reasons.

8.2 Assessment plan

Year mark Final mark

8.3 Assignment numbers

8.3.2 Unique assignment numbers

Consider the set

Consider the set