Mathematics 206

Solutions for HWK 17a

Section 5.4

Section 5.4 p244 Problem 1. In each part, explain why the given vectors do not form a basis
for the indicated vector space. (Solve the problem by inspection.)

(a) u1 = (1, 2), u2 = (0, 3), u3 = (2, 7) for R2

(b) u1 = (−1, 3, 2), u2 = (6, 1, 1) for R3

(c) p1 = 1 + x + x2 , p2 = x − 1 for P2 .
· ¸ · ¸ · ¸ · ¸ · ¸
1 1 6 0 3 0 5 1 7 1
(d) A = , B= , C= , D= , E= , for M22
2 3 −1 4 1 7 4 2 2 9

Solution.

(a) Too many vectors: 3 vectors in the 2-dimensional space R2 must form a dependent set.

(b) Too few vectors: 2 vectors in the 3-dimensional space R3 cannot span R3 .

(c) Too few: 2 vectors in the 3-dimensional space P2 cannot span P2 .

(d) Too many: 5 vectors in the 4-dimensional space M22 must form a dependent set.

Section 5.4 p244 Problem 3b. Do the vectors (3, 1, −4), (2, 5, 6), (1, 4, 8) form a basis for R3 ?

Solution. Since we have the correct count (3 vectors for a 3-dimensional space) there is certainly
a chance. If these 3 vectors form an independent set, then one of the theorems in 5.4 tells us that
they’ll form a basis. If not, they can’t form a basis. So we need only test for independence. So:
does the equation k1 (3, 1, −4) + k2 (2, 5, 6) + k3 (1, 4, 8) = (0, 0, 0) have nontrivial solutions or not?
Equivalently, does the system
    
3 2 1 k1 0
 1 5 4   k2  =  0 
−4 6 8 k3 0

have nontrivial solutions? According to the Long Theorem, the answer depends on whether the
indicated coefficient matrix has determinant 0 or not. Since
 
3 2 1
det  1 5 4  = 42 6= 0
−4 6 8

there are no nontrivial solutions. The given vectors form an independent set, hence a basis for R3 .

Page 1 of 8 A. Sontag April 11, 2002

 Math 206 HWK 17a Solns contd
5.4 p244

Section 5.4 p244 Problem 4b. Do the vectors/functions 4 + 6x + x2 , −1 + 4x + 2x2 , 5 + 2x − x2

form a basis for P2 ?

Solution. Once again we have the correct count, since P2 has dimension 3. We need only test
for independence. So consider the condition

k1 (4 + 6x + x2 ) + k2 (−1 + 4x + 2x2 ) + k3 (5 + 2x − x2 ) = 0

for every x. This is equivalent to the system

    
4 −1 5 k1 0
6 4 2   k2  =  0 
1 2 −1 k3 0

This time we find  

4 −1 5
det  6 4 2 =0
1 2 −1

According to the Long Theorem, we have a homogeneous system with nontrivial solutions. There-
fore the given three vectors/functions form a dependent set, hence do not form a basis.

Section 5.4 p244 Problem 5. Show that the set {A, B, C, D} is a basis for M22 .
· ¸ · ¸ · ¸ · ¸
3 6 0 −1 0 −8 1 0
A= , B= , C= , D=
3 −6 −1 0 −12 −4 −1 2

Solution. (or outline of solution). It suffices to show that the set {A,·B, C, ¸
D} is independent.
0 0
(Why?) To this end, consider the equation k1 A + k2 B + k3 C + k4 D = . This equation is
0 0
equivalent to the system
    
3 0 0 1 k1 0
 6 −1 −8 0   k2   0 
   =  .
3 −1 −12 −1 k3 0
−6 0 −4 2 k4 0

One can show that this system has no nontrivial solutions. (Do this.) Therefore the only values of
k1 , . . . k4 that will give us · ¸
0 0
k1 A + k2 B + k3 C + k4 D =
0 0

Page 2 of 8 A. Sontag April 11, 2002

 Math 206 HWK 17a Solns contd
5.4 p244

are
k1 = k2 = k3 = k4 = 0.
This finishes the proof. (Why?)

Section 5.4 p244 Problem 12. Find a basis for and the dimension of the solution space for
the homogeneous system
3x1 + x2 + x3 + x4 = 0
5x1 − x2 + x3 − x4 = 0

Solution. First solve the given system, using row reduction. The augmented matrix for the
given system is · ¸
3 1 1 1 | 0
5 −1 1 −1 | 0
which reduces to · ¸
1
1 0 4 0 | 0
1 .
0 1 4
1 | 0

The nonleading variables are x3 and x4 . Setting x3 = s and x4 = t, and solving for the leading
variables in terms of the nonleading ones, we find a parametric description for the solutions. The
solutions are of the form
1 1
(x1 , x2 , x3 , x4 ) = s(− , − , 1, 0, ) + t(0, −1, 0, 1).
4 4

We can read off from this description that the vectors

(−1, −1, 4, 0), (0, −1, 0, 1)

span the solution space. By inspection, we can see that they form an independent set (why?).
Therefore the set
{(−1, −1, 4, 0), (0, −1, 0, 1)}

is a basis for the solution space, which has dimension 2. Other correct answers are possible.

Page 3 of 8 A. Sontag April 11, 2002

 Math 206 HWK 17a Solns contd
5.4 p244

Section 5.4 p244 Problem 17. Find bases for the following subspaces of R3 .

(a) The plane 3x − 2y + 5z = 0

(b) The plane x − y = 0

(c) The line x = 2t, y = −t, z = 4t

(d) The set of all vectors of the form (a, b, c) where b = a + c

Solution. (a) Find a basis for the solution space to the equation 3x − 2y + 5z = 0. First find a
parametric description for the solutions. We can do this by putting y = s, z = t, and solving for
x. The solutions are (x, y, z) = s( 23 , 1, 0) + t(− 53 , 0, 1). Therefore the vectors

(2, 3, 0), (−5, 0, 3)

span the solution space. They also form an independent set, so they form a basis. Since there are
exactly two vectors in the basis, the dimension of the solution space is 2. In other words, as we
knew all along, the dimension of the specified plane is 2.

(b) One method is to follow the same procedure as for (a). The solutions for the equation x − y = 0
can be described by
(x, y, z) = s(1, 1, 0) + t(0, 0, 1).

The vectors (1, 1, 0) and (0, 0, 1) span the solution set for x − y = 0 and they form an independent
set. Hence they form a basis for the plane x − y = 0, a 2-dimensional subspace of R3 .

Here’s another method. The points on the given plane are exactly the points of the form (x, x, z)
where x and z are real numbers. In other words, they’re the points of the form x(1, 1, 0)+ z(0, 0, 1).
Now read off that the set
{(1, 1, 0), (0, 0, 1)}

is a basis for the plane in question.

(c) Points on this line are of the form t(2, −1, 4) where t can be any real number. Thus the vector
(2, −1, 4) spans this line. Moreover, the singleton set

{(2, −1, 4)}

is independent. Thus it forms a basis for the line in question, which therefore (as we anticipated)
has dimension 1.

Page 4 of 8 A. Sontag April 11, 2002

 Math 206 HWK 17a Solns contd
5.4 p244

(d) The vectors in question can also be written in the form (a, a+c, c), hence in the form a(1, 1, 0)+
c(0, 1, 1). Together, the vectors
(1, 1, 0), (0, 1, 1)

span the space in question. They form an independent set, hence a basis. The set in question has
dimension 2.

Section 5.4 p244 Problem 18. Find the dimensions of the following subspaces of R4 .

(a) The set of all vectors of the form (a, b, c, 0).

(b) The set of all vectors of the form (a, b, c, d) where d = a + b and c = a − b.

(c) The set of all vectors of the form (a, b, c, d) where a = b = c = d.

Solution. (a) Let W be the subspace in question. The vectors in W can be written in the form
a(1, 0, 0, 0) + b(0, 1, 0, 0) + c(0, 0, 1, 0). By inspection, we see that {(1, 0, 0, 0), (0, 1, 0, 0), (0, 0, 1, 0)}
forms a basis for W . Therefore W has dimension 3.

(b) Denote the given subspace by W . Vectors in W can be written in the form (a, b, a − b, a + b),
hence in the form a(1, 1, 1, 0) + b(0, 1, −1, 1). This time we see that the set {(1, 1, 1, 0), (0, 1, −1, 1)}
is a basis for W , which therefore has dimension 2.

(c) Denote the subspace by W . Vectors in W are those of the form (a, a, a, a), hence of the form
a(1, 1, 1, 1). The singleton set {(1, 1, 1, 1)} forms a basis for W , which is therefore a 1-dimensional
subspace of R4 .

Section 5.4 p244 Problem 21. Find standard basis vectors that can be added to the set
4
{v1 , v2 } to produce a basis for R , given that

v1 = (1, −4, 2, −3), v2 = (−3, 8, −4, 6)

Solution. Here’s one solution. The standard basis vector (0, 0, 0, 1), which I’ll call u, is not
a linear combination of v1 and v2 (verify this). According to the Plus/Minus Theorem, the set
{v1 , v2 , u} is independent. Inspecting the second and third entries in v1 and v2 it looks like we
could use either (0, 1, 0, 0) or (0, 0, 1, 0) for the fourth vector. Pick one (either one will work) and
test the new collection of four vectors to confirm independence. Other correct answers are possible.
(See the text for a list of all possible correct answers.)

Page 5 of 8 A. Sontag April 11, 2002

 Math 206 HWK 17a Solns contd
5.4 p244

Note: I’d hoped that Problem 21 could serve as a lead-in to our proof that independent sets with
too few vectors to form a basis can be expanded until they become a basis. It didn’t work too well
for that. Various other solutions may well work better than what’s given here.

Section 5.4 p244 Problem 22. This now appears on GP19. A solution will be provided later.
In the meanwhile, you might like to consult the solution outline that’s given here for Problem 5.

Section 5.4 p244 Problem 23ab.

(a) Show that for every positive integer n, one can find n + 1 linearly independent vectors in
F (∞, ∞). Hint: Look for polynomials.

(b) Use the result in part (a) to prove that F (−∞, ∞) is infinite-dimensional.

Solution. (a) Given a positive integer n, consider the set {1, x, x2 , . . . , xn }. We already know
that this set is independent. Obviously there are n + 1 vectors/functions in this set. Done.

(b) Part (a) shows that there is no upper bound on the size of an independent set in this space. If
this space were finite dimensional, then it would have a finite basis and any set with more vectors
than the number of vectors in the basis would have to be dependent. This impossibility shows that
the space can’t have finite dimension.

Section 5.4 p244 Problem 28ab. The figure given in the text shows a rectangular xy-
coordinate system determined by the unit basis vectors i and j and an x0 y 0 -coordinate system
determined by unit basis vectors u1 and u2 . Find the x0 y 0 -coordinates of the points whose xy-
coordinates are given.

◦
In the figure, u2 = j and u1 is a unit vector that √ makes an angle of 30 with the positive x-axis.
3 1
Thus u2 = (0, 1) and u1 = (cos 30◦ , sin 30◦ ) = ( , ).
2 2
√
(a) ( 3, 1) (b) (1, 0)

√
Solution. (a) By inspection, ( 3, 1) = 2u1 + 0u2 . The x0 y 0 -coordinate vector is (2, 0).
√
3 1 2 a 1
(b) Solve the system (1, 0) = a( , ) + b(0, 1) to find a = √ , b = − = − √ . Thus (1, 0) =
2 2 3 2 3

Page 6 of 8 A. Sontag April 11, 2002

 Math 206 HWK 17a Solns contd
5.4 p244

√
2 3 1 1 2 1
√ ( , ) − √ (0, 1)+. The x0 y 0 -coordinate vector is ( √ , − √ ).
3 2 2 3 3 3

Section 5.4 p244 Problem 31. Find the dimension for each of the following vector spaces.

(a) The vector space of all diagonal n × n matrices.

(b) The vector space of all symmetric n × n matrices.

(c) The vector space of all upper triangular n × n matrices.

Solution. (a) Let Dnn denote the specified vector space of diagonal matrices. One basis for Dnn
consists of the n different n × n matrices, each of which has exactly one of the diagonal entries
equal to 1 and all other entries in the matrix equal to 0. [For D33 , for instance, this basis would
use the 3 matrices      
1 0 0 0 0 0 0 0 0
0 0 0, 0 1 0, 0 0 0
0 0 0 0 0 0 0 0 1
as basis vectors.
Thus Dnn has dimension n.

(b) Let Snn denote the specified vector space of symmetric matrices. Let’s look at a special case
first, just to get oriented. Consider S33 for instance. For a particular 3 × 3 matrix to belong to
Snn the entries on the main diagonal can be anything, but the entries off the main diagonal must
be paired up to match symmetrically. A typical matrix in S33 would look like
 
a x y
x b z
y z c

One basis for S33 would use the following 6 matrices as the basis vectors:
           
1 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0
0 0 0, 0 1 0, 0 0 0, 1 0 0, 0 0 0, 0 0 1
0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 1 0

So S33 would have dimension 6.

Page 7 of 8 A. Sontag April 11, 2002

 Math 206 HWK 17a Solns contd
5.4 p244

More generally to create a basis for Snn we could use the following:
• n diagonal matrices, each having a single 1 on the diagonal and all other diagonal entries 0
• n − 1 matrices, each having exactly one first-row entry above the diagonal equalling 1, the
symmetric entry equalling 1, and all other entries equalling 0,
• n − 2 matrices, each having exactly one second-row entry above the diagonal equalling 1,
the symmetric entry equalling 1, and all other entries equalling 0,
• and so forth, continuing through
• 1 matrix having the very last entry of the next to last row equalling 1, the symmetric entry
in the last row equalling 1, and all other entries equalling 0.

This gives
n + (n − 1) + (n − 2) + . . . + (n − (n − 1))

matrices in the basis. The total number of basis vectors is therefore

n(n − 1) n2 + n n(n + 1)
n2 − (1 + 2 + . . . + (n − 1)) = n2 − = = .
2 2 2

The combinatorialists among you may well find a faster way to count these. In any case,
n(n + 1)
dim(Snn ) = .
2
n(n + 1)
(c) A variation on the reasoning used in (b) gives the dimension to be . Just leave out
2
the bit about making the relevant symmetric entry equal to 1.

Section 5.4 p244 Problem 33. State the two parts of Theorem 5.4.2 in contrapositive form.

Solution. Theorem 5.4.2 has the underlying assumption that V is a finite-dimensional vector
space with {v1 , v2 , . . . , vn } a basis for V . Under this assumption, the two parts of the theorem
read as follows:

(a) If a set has more than n vectors, then it is linearly dependent.

(b) If a set has fewer than n vectors, then it does not span V .

In contrapositive form, these become:

(a1) If a set is not linearly dependent, then it does not have more than n vectors.
(b1) If a set does not not span V then it does not have fewer than n vectors.

Each of these contrapositives would read better if reworded. Here are possible rewordings.
(a2) If a set is linearly independent, then it has at most n vectors.
(b2) If a set spans V , then it has at least n vectors.

Page 8 of 8 A. Sontag April 11, 2002