Linear Algebra

Exercise sheet 6

1. Which of the following sets of vectors are bases? (motivate the answer)
           
 1 1   1 1 1 1 
A=  1 , 0
   B=  2 , 3 , 0 , 0
     
1 1 3 5 1 5
   

           
 1 1 2   1 1 5 
C=  1 , 2 , −1
     D=  1 , 2 , 3
   
1 3 1 2 5 4
   

Solution: A basis of R3 must have exactly 3 elements so A, B are not bases , C is a basis since the
matrix whose columns are the three vectors has nonzero determinant, D is not a basis since the matrix has
determinant zero.

2. Given the vectors v 1 = (1, 2, 3), v 2 = (1, 1, 1), v 3 = (0, 1, 1), v 4 = (0, 1, 0)
(a) Check they span R3 and then extract a basis from them.
(b) Write a linear combination a1 v 1 + a2 v 2 + a3 v 3 + a4 v 4 = 0 with at least one coefficient ai 6= 0
(c) Write (1, 1, 0) as a linear combination of the vectors v 1 , v 2 , v 3 , v 4 in two different ways.

Solution: For (a) we take any three such that the matrix whose columns are the chosen vectors has nonzero
determinant. For example, we can choose v 2 , v 3 , v 4 .
(b) Since the chosen vectors are a basis of R3 , we can write the other one as a linear combination of the
others. With the choice made above, v 1 = v 2 + 2v 3 − v 4 , so that 0 = v 1 − v 2 − 2v 3 + v 4 .
(c) Write (1, 0, 0) as a linear combination of the chosen vectors plus 0 times the other vector. This is the
first way. To get the second add the linear combination giving 0 found in (b). E.g.

(1, 1, 0) = 0v 1 + v 2 − v 3 + v 4 (1, 1, 0) = 0v 1 + v 2 − v 3 + v 4 + v 1 − v 2 − 2v 3 + v 4 = v 1 − 3v 3 + 2v 4 .

3. Let U ⊆ R4 be the subspace spanned by the vectors v 1 = (1, 2, 2, 0), v 2 = (1, 0, 0, 1)v 3 =
(0, 4, 2, 0), v 4 = (1, 2, 1, 1).
(a) Find the dimension of U
(b) Find two different (not just two different orderings of the same basis) bases of U . Si scriva
la matrice di passaggio tra queste basi

Solution: We have v 4 = v 2 + 12 v 3 , so the dimension of U is 3 since v 1 , v 2 , v 3 are linearly independent.

Any basis of U must contain v 1 since it can’t be written as a linear combination of the other vectors , a
possible choice is v 1 , v 2 , v 4 .

4. Let R4 [x] be the space of polynomials of degree strictly less than 4 Considere the subsets

U = p(x) ∈ R4 [x]|p(2) = 0 V = p(x) ∈ R4 [x]|p(2) = p(0) = 0 W = p(x) ∈ R4 [x]|p0 (2) = 0

  

Where p0 (x) is the first derivative of p(x). Check these are subspaces, find the dimension and a
basis for each of them.
 Solution: That the three are subspaces is an easy check. Now if p(x) = a + bx + cx2 + dx3 and p(2) = 0,
then we have the linear equation in the coefficients a + 2b + 4c + 8d = 0, so the dimension of U is 3 and a
basis is x − 2, x2 − 4, x3 − 8. We could also have noticed that x − 2, (x − 2)2 , (x − 2)3 are linearly independent
and the dimension of U can not be 4 since U is not the whole space (the monomial 1 does not belong to it).
So dimension is 3 and also x − 2, (x − 2)2 , (x − 2)3 form a basis of U . In the same way we can note that all
polynomials in V must have degree bigger than 1, and that V is a subspace of U , hence (x − 2)x, (x − 2)x2
is a basis. Finally, coefficients of the polynomials in W must satisfy the equation b + 4c + 12d = 0 so the
dimension is 3, and a basis is given by 1, x2 − 4x, x3 − 12x.

5. Find the dimension and a basis for the space generated by the rows (the row space) and by the
columns (column space) of the following matrices.
   
1 1 1 1 2 0 −1
A = 2 2 3 , B = 2 6 −3 −3
1 0 1 3 10 −6 −5

Solution: The determinant of A is 1, so the columns are linearly independent and the dimension is 3.
Also the rows are all linearly independent: we can see that either by observing that the determinant of At
is also 1 or that A is row equivalent to the identity matrix by the algorithm used to compute A−1 using
Gaussian reduction. The matrix B is row equivalent to the row echelon matrix
 
1 2 0 −1
B 0 = 0 2 −3 −1
0 0 0 0

By an exercise in the previous sheet, row equivalent matrices have the same row space, hence the dimension
of the row space is 2 and a basis is given by the first two rows of either B or its reduction. For the column
space we notice that a parametric solution to Bx = 0 is given by
       
1 2 0 −1
s + 3t  
−3t 2 + 6 + t −3 + s −3
2
3 10 −6 −5
from which we get (putting s = 1, t = 0 and s = 0, t = 1) that the third and fourth columns of B are linear
combinations of the first two which are independent and form a basis of the column space.

6. Given the vectors L = v 1 = (3, 1, 2)t , v 2 = (2, 1, 3)t find the cartesian equations for the space
generated by them and add a vector to get a basis of R3 .

Solution: Consider the matrix Mat(v 1 , v 2 , x) where x is a vector of unknowns. Its determinant is zero
if and only if x ∈ L[v 1 , v 2 ] so computing it and equating the result to zero we get the cartesian equation
x − 5y + z = 0. Any vector u not satisfying this equation does not belong to L[v 1 , v 2 ] and so v 1 , v 2 , u are
linearly independent and form a basis of R3 .

7. Given the vectors v 1 = (1, 1, 1, 1)t , v 2 = (2, 2, 3, 4)t find cartesian equations for the space they
generate, and then add two vectors to get a basis of R4 .

Solution: The generic linear combination of the two vectors has components x1 = s + 2t, x2 = s + 2t, x3 =
s + 3t, x4 = s + 4t. A system of cartesian equations for a codimension two subspace has two equations.
We see immediately the two vectors verify x1 = x2 , so we can take that as a first equation. We also
have x4 − x3 = t, x3 − x2 = t so we can take as second equation x2 − 2x3 + x4 = 0. Now the vector
e4 = (0, 0, 0, 1)t does not satisfy the second equation, so it does not belong to L[v 1 , v 2 ] and the vectors
ulv1 , v 2 , e4 are linearly independent. We can find the equation for the space generated by these vectors by
computing the determinant of Mat(v 1 , v 2 , e4 , x), where x is a vector of unknowns. We get (up to a sign)
x1 −x2 , so x1 −x2 = 0 is a cartesian equation for L[v 1 , v 2 , e4 ], hence ulv1 , v 2 , e4 , e1 are linearly independent
and form a basis of R4
8. Given the following bases of R3 :
                 
 1 0 0   2 −2 1   1 1 1 
C = 0 , 1 , 0 B1 = 2 ,  1  , −2 B2 = 1 , 1 , 0
0 0 1 1 2 2 1 0 0
     

denote by A1 , A2 the matrices whose columns are the vectors of B1 and of B2

(a) Check that the coordinates of the vectors of the canonical basis with respect to B1 , B2 are
given by the columns of the matrices A−1 −1
1 , A2 .
(b) Find the coordinates of the vector (a, b, c)t with respect to the bases B1 , B2 .

Solution: The first part is just a verification. For the second part, denote by a the vector (a, b, c)t and by
u1 , u2 , u3 the vectors of B1 , we have

a = (e1 , e2 , e3 )a = (u1 , u2 , u3 )A−1

1 a

So the coordinates of a w.r.t. B1 are A−1

1 a

9. Suppose U ⊆ V is a subspace of dimension n = dim (V ). Show U = V .

Solution: Let v ∈ V, {u1 , . . . , un } a basis of U . Since the dimension of V is n, any basis of V is a spanning
set with n elements, so the vectors u1 , . . . , un v are linearly dependent. Let a1 u1 + . . . + an un + bv = 0 be a
linear combination to 0 with at least one nonzero coefficient. Since the ui ’s are linearly independent b can
not be zero (why?) so v is a linear combination of the vectors ui ’s and hence belongs to U .