Z

 ZZ
 Exo7 
Z 
Z
Z

Année 2009
Exercices de mathématiques

Exercises and corrections: Barbara Tumpach

Linear maps – subvectorspaces of Rn
Exercise 1. 1. Endow R2 with an orthonormal frame (O,~i, ~j). Show that
a linear map from R2 to R2 is uniquely determined by its values on the
vectors ~i and ~j.

2. In the basis {~i, ~j}, what is the matrix of the orthogonal symmetry with
respect to the horizontal axis?

3. In the basis {~i, ~j}, what is the matrix of the orthogonal projection to
the horizontal axis?

4. In the basis {~i, ~j}, what is the matrix of the rotation of angle θ and
center O?

5. In the basis {~i, ~j}, what is the matrix of the homothety of center O
and ratio k?

6. In the basis {~i, ~j}, what is the matrix of the symmetry of center O?

7. Is a translation a linear map?

Correction 1. 1. Let f : R2 → R2 be a linear map. Consider any vector

~v in R . Since {~i, ~j} is a basis of R2 , ~v can be uniquely written as : ~v =
2

x~i + y~j. By linearity of f , one has : f (~v ) = f (x~i + y~j) = xf (~i) + yf (~j).
Therefore the values of f on the vectors ~i and ~j, determine the value
of f on any vector of R2 . Two linear maps taking the same values on
~i and ~j will coincide on R2 .

2. In the basis {~i, ~j}, the matrix

 of the
 orthogonal symmetry with respect
1 0
to the horizontal axis is .
0 −1

3. In the basis {~i,~j}, thematrix of the orthogonal projection to the hori-

1 0
zontal axis is .
0 0

1
 4. In the basis {~i, ~j}, 
the matrix of the rotation of angle θ and center O
cos θ − sin θ
is .
sin θ cos θ

5. In the
 basis {~i, ~j}, the matrix of the homothety of center O and ratio
k 0
k is .
0 k
 
~ ~ −1 0
6. In the basis {i, j}, the matrix of the symmetry of center O is .
0 −1

7. A linear map f from Rn into Rp necessarily maps ~0 ∈ Rn onto ~0 ∈ Rp .

The translation by a given vector ~u ∈ R2 takes ~v ∈ R2 to ~v + ~u ∈ R2 .
In particular, the translation of vector ~u takes ~0 ∈ R2 to ~u ∈ R2 .
Therefore, if ~u 6= ~0, the translation of vector ~u is not a linear map.

Exercise 2. Let f be the map from R4 to R4 defined by:

f (x, y, z, t) = (x + y + z + t, x + y + z + t, x + y + z + t, 2x + 2y + 2z + 2t) .

1. Show that f is linear and write down its matrix in the canonical basis
of R4 .

2. Check that the vectors ~a = (1, −1, 0, 0), ~b = (0, 1, −1, 0) and ~c =
(0, 0, 1, −1) all belong to ker f .

3. Check that the vector d~ = (5, 5, 5, 10) belongs to Imf .

Correction 2. 1. One has to verify that, for any vectors v~1 and v~2 in
4
R and any λ ∈ R, one has f (v~1 + λv~2 ) = f (v~1 ) + λf (v~2 ). Denote
by (x1 , y1 , z1 , t1 ) (resp. (x2 , y2 , z2 , t2 )) the coordinates of the vector v~1
(resp. v~2 ) in the canonical basis of R4 . The coordinates of the vector
v~1 + λv~2 are (x1 + λx2 , y1 + λy2 , z1 + λz2 , t1 + λt2 ). Therefore, using the
formula that defines the map f , one has:

f (v~1 + λv~2 ) = f (x1 + λx2 , y1 + λy2 , z1 + λz2 , t1 + λt2 )

 
x1 + λx2 + y1 + λy2 + z1 + λz2 + t1 + λt2
 x1 + λx2 + y1 + λy2 + z1 + λz2 + t1 + λt2 
= .
 x1 + λx2 + y1 + λy2 + z1 + λz2 + t1 + λt2 
2(x1 + λx2 ) + 2(y1 + λy2 ) + 2(z1 + λz2 ) + 2(t1 + λt2 )

2
 On the other hand,
   
x1 + y1 + z1 + t1 x2 + y2 + z2 + t2
 x1 + y1 + z1 + t1   x2 + y2 + z2 + t2 
f (v~1 ) = 
 x1 + y1 + z1 + t1  ; f (v~2 ) =  x2 + y2 + z2 + t2
  ;

2x1 + 2y1 + 2z1 + 2t1 2x2 + 2y2 + 2z2 + 2t2
 
λ(x2 + y2 + z2 + t2 )
 λ(x2 + y2 + z2 + t2 ) 
λf (v~2 ) = 
 λ(x2 + y2 + z2 + t2 )  ;


λ(2x2 + 2y2 + 2z2 + 2t2 )

and
 
x1 + y1 + z1 + t1 + λ(x2 + y2 + z2 + t2 )
 x1 + y1 + z1 + t1 + λ(x2 + y2 + z2 + t2 ) 
f (v~1 )+λf (v~2 ) =  .
 x1 + y1 + z1 + t1 + λ(x2 + y2 + z2 + t2 ) 
2x2 + 2y2 + 2z2 + 2t2 + λ(2x2 + 2y2 + 2z2 + 2t2 )

By commutativity of the reals, one obtains f (v~1 +λv~2 ) = f (v~1 )+λf (v~2 ).
The matrix of f in the canonical basis of R4 is
 
1 1 1 1
 1 1 1 1 
 1 1 1 1 .
 

2 2 2 2

2. Let us compute the images of the vectors ~a = (1, −1, 0, 0), ~b = (0, 1, −1, 0)
and ~c = (0, 0, 1, −1). One has

f (~a) = f (1, −1, 0, 0) = (1 − 1, 1 − 1, 1 − 1, 2 − 2) = (0, 0, 0, 0);

f (~b) = f (0, 1, −1, 0) = (1 − 1, 1 − 1, 1 − 1, 2 − 2) = (0, 0, 0, 0);

f (~c) = f (0, 0, 1, −1) = (1 − 1, 1 − 1, 1 − 1, 2 − 2) = (0, 0, 0, 0).
Therefore ~a, ~b and ~c belong to kerf .

3. Since the vector d~ = (5, 5, 5, 10) is the image of the vector (5, 0, 0, 0), d~
belongs to Imf .

Exercise 3. Consider the map f : R3 → R3 given by:

f (x, y, z) = (x + 2y + z, 2x + y + 3z, −x − y − z).

3
 1. Justify that f is linear.

2. Give the matrix of f in the canonical basis of R3 .

3. (a) Determine a basis and the dimension of the kernel of f , denoted

by kerf .
(b) Is the map f injective?

4. (a) Give the rank of f and a basis of Imf .

(b) Is the map f surjective?

Correction 3. 1. One has to verify that, for any vectors v~1 and v~2 in R3
and any λ ∈ R, one has f (v~1 + λv~2 ) = f (v~1 ) + λf (v~2 ). It is the same
kind of computation as in Exercise 2, question 1.

2. The matrix of f in the canonical basis of R3 is

 
1 2 1
 2 1 3 .
−1 −1 −1

3. (a) The kernel of f , written ker f , is the set of vectors which are
mapped onto ~0 by f . Therefore, a vector ~v = (x, y, z) ∈ R3
belongs to kerf if and only if (x, y, z) is a solution of the following
system: 
 x + 2y + z = 0
2x + y + 3z = 0
−x − y − z = 0


Applying the Gauss method, one obtains that the above system
is equivalent to
 
 x + 2y + z = 0  x + 2y + z = 0
⇔ −3y + 2z = 0 ⇔ −3y + 2z = 0 .
−3y − 2z = 0 4z = 0
 

Therefore the unique solution of the system is ~v = ~0, and kerf =

{~0}. The dimension of kerf is therefore 0. The empty set ∅ is a
basis of kerf .
(b) For a linear map, being injective is equivalent to kerf = {~0}.
Hence, by the previous question, f is injective.

4
4. (a) There are many ways to answer this question. Recall that a vector
~b ∈ R3 belongs to Imf if and only if there exists ~v = (x, y, z) ∈ R3
such that f (~v ) = ~b, or equivalently if ~b is a linear combination
of the columns of the matrix associated to f . According to the
expression of the matrix associated to f given in question  2.,

1
Imf is the vector space generated by the vectors C1 =  2 ,
    −1
2 1
C2 =  1  , C3 =  3 .
−1 −1
To find a basis of Imf , one way is to apply Gauss algorithm to
the matrix of f in order to trigonalize it. One finds:
C2 ← C2 − 2C1
 C1 C2 C3  C3 ← 3C3 + C2
C3 ← C3 − C1

1 2 1 1 0 0
→ 1 0 0 →  .
 2 1 3   2 −3 1  2 −3 0 
−1 −1 −1 −1 1 1
−1 1 0
     
1 0 0
Since the vectors  2  ,  −3 , and
  0  are column vec-
−1 1 1
tors of a triangular matrix, they are linearly independent. Since
we applied the Gauss algorithm to the columns of the matrix as-
sociated to f , they generate Imf . Consequently they form a basis
of Imf which is therefore equal to R3 .
Another way to find a basis of Imf , is to compute the determinant
of the matrix associated to f . Since
 
1 2 1
det  2 1 3  6= 0,
−1 −1 −1
the columns of this matrix are linearly independent. Therefore
they form a basis of Imf .
A shorter way to answer this question, is to use Rank Theorem.
Since f is an injective map from R3 into R3 , one has
dim R3 = dim kerf + dim Imf ⇔ 3 = 0 + dim Imf.
Hence Imf = R3 since the only subspace of dimension 3 of R3 is R3
itself. One concludes that the rank of f (which is by definition the

5
 dimension
 ofImf ) is3, and
 a basis ofImfis given, for example,
1 0 0
~
by i =  0 ,j=
 ~   ~
1 , and k =  0 .
0 0 1
(b) Recall that a map f : E → F is surjective if and only if Imf = F .
For a linear map, this is equivalent to dim Imf = dim F . By the
previous question, the map considered in this exercise is surjective.

Exercise 4. 1. Let f be a surjective linear map from R4 to R2 . What is

the dimension of the kernel of f ?

2. Let g be an injective map from R26 to R100 . What is the dimension of

the image of g?

3. Can there be a bijective linear map from R50 to R72 ?

Correction 4. 1. By the Rank Theorem, dim ker f = dim R4 −dim Im f .

Since f is supposed to be surjective, dim Imf = 2. Therefore dim ker f =
4 − 2 = 2.

2. By the Rank Theorem, dim Im g = dim R26 − dim ker g. Since g is

supposed to be injective, dim ker g = 0. Hence dim Im g = 26.

3. By the Rank Theorem, an injective map from R50 to R72 satisfies

dim Im f = 50. On the other hand, a surjective map from R50 to
R72 satisfies dim Im f = 72. Consequently a map from R50 to R72
can not be injective and surjective. Therefore there exists no bijective
maps from R50 to R72 .

Exercise 5. Consider the matrix

 
2 7 1
A =  −1 2 0  .
3 5 1

1. Compute a basis of the kernel of A.

2. Compute a basis of the image of A.

Correction 5. We will answer both questions at the same time. To do so,

we will apply the Gauss algorithm to the columns of the matrix A and I3

6
simultanously (here I3 denotes the identity matrix of size (3, 3) having the
same number of columns as A).

C2 ← 2C2 − 7C1
   C3 ← 2C3 − C1  C3 ← 11C3 − C2
2 7 1 2 0 0 2 0 0
A=  −1 2 0   −1 11 1   −1 11 0 
→ → .
 3 5 1  3 −11 −1  3 −11 0 
1 0 0 1 −7 −1 1 −7 −4
I3 =  0 1 0   0 2 0   0 2 −2 
0 0 1 0 0 2 0 0 22
 
2
It follows that a basis of Im A is given by the vectors v~1 =  −1  and
  3
0
v~2 =  11 . Indeed, the columns of the upper matrix still generate Im A
−11
since we obtained them by applying the Gauss algorithm to the columns of
A. The third column of the upper matrix being equal to the null vector, we
only consider the first two columns, namely the vectors v~1 and v~2 . These two
vectors are linearly independant since they are two columns of a triangular
matrix.  
−4
On the other hand the kernel of A is generated by the vector ~u =  −2 .
22
3
Indeed, by the Rank Theorem dim ker f = dim R − dim Im f = 1 since
dim Im f = 2. Moreover, one can verify that ~u is a non-zero vector of ker f
by:     
2 7 1 −4 −8 − 14 + 22 = 0
A ~u =  −1 2 0   −2  =  4−4=0 .
3 5 1 22 −12 − 10 + 22 = 0
Exercise 6. Consider the matrix
 
1 2 3 1
B =  −1 2 −1 −3  .
−3 5 2 −3

1. Compute a basis of the kernel of B.

2. Compute a basis of the image of B.

7
Correction 6. We use the same technique as in the previous exercise: the
Gauss algorithm on the columns of the matrix B and I4 simultaneously (here
I4 denotes the identity matrix having as many columns as B, namely 4
columns).
C2 ← C2 − 2C1
C3 ← C3 − 3C1
  C4 − C1
1 2 3 1  
1 0 0 0
B =  −1 2 −1 −3   −1 4 2 −2 
−3 5 2 −3 →
  −3 11 11 0
1 0 0 0  
 0 1 0 0  1 −2 −3 −1
I4 = 
 0 0 1 0 
 
 0 1 0 0 

 0 0 1 0 
0 0 0 1
0 0 0 1

C3 ← 2C3 − C2
C4 ← 2C4 + C2  C4 ← C4 − C3 
  1 0 0 0
1 0 0 0  −1 4 0 0 
 −1 4 0 0 
→  −3 11 11 11  −3 11 11 0 
 → 1 −2 −4 0
1 −2 −4 −4  0 1 −1 2 
 0 1 −1 1   
   0 0 2 −2 
 0 0 2 0 
0 0 0 2
0 0 0 2
 
1
Consequently, a basis of Im f is given by the three vectors v~1 =  −1 ,
    −3
0 0
v~2 =  4  and v~3 =  0 . A basis of ker f is given by the vector
11 11
 
0
 2 
~u =  −2 .


2
Exercise 7. Consider the matrix
 
−1 3 1
 1 2 0 
 
C=  2 −1 −1  .

 2 4 0 
1 7 1
1. Compute a basis of the kernel of C.
2. Compute a basis of the image of C.

8
Correction 7. One has
C2 ← C2 + 3C1
C3 ← 5C3 − C
2
C3 ← C3 + C
1 
  −1 0 0
−1 3 1 −1 0 0  1 5 0 
 1 2 0   1 5 1   
     2 5 0 
C =  2 −1 −1 
 
 2 5 1  →  2 10 0  .
   
 2 →
4 0   2 10 2 
 1 10 0 
1 7 1  1 10 2 1 3 2
1 0 0 1 3 1  0 1 −1 
I3 =  0 1 0   0 1 0 
0 0 5
0 0 1 0 0 1
 
−1

 1 

Consequently a basis of Im f is given by the two vectors v~1 = 
 2  and

 2 
1
 
0  

 5 
 2
v~2 =  5 . A basis of ker f is given by the vector ~u =  −1 .
 
 10  5
10