Partial Solution Set, Leon Section 4.

4.2.2a Define L : R3 R2 by L (x1 , x2 , x3 )T = (x1 + x2 , 0)T . Find a matrix A such that

L(x) = Ax for every x R3 .

Solution: We take as the columns of
 A the images under L of the standard basis vectors
1 1 0
from R2 , obtaining A = .
0 0 0

4.2.4 L : R3 R3 is given by L(x) = (2x1 x2 x3 , 2x2 x1 x3 , 2x3 x1 x2 )T . The

standard matrix representation of L is (again) the matrix A whose ith column is L(ei ),
i.e.,

2 1 1
A = 1 2 1 .
1 1 2

4.2.5b L is the operator on R2 whose effect is to first reflect a vector x about the x1 axis and
then rotate it counterclockwise through an angle of /2. We are to find the standard
matrix representation for L.
This can be found by considering the effect of the transformation on the standard basis
vectors: e1 is unaffected by the reflection, and the subsequent rotation produces a copy
of e2 , while the reflection sends e2 to e2 , and the subsequent rotation sends e
 2 to e1 .
0 1
The overall effect of L is reflection about the identity line. The matrix is A = .
1 0
Note that we can do this in a different (and simpler, if messier) way. It is not hard to show
that the matrix representation of the composition of transformations is the product of the
individual matrix representations. So let L = L2 L1 , where L1 is the reflection  and L2
1 0
is the rotation. The standard matrix representations for L1 and L2 are A1 =
  0 1
0 1
and A2 = . Thus
1 0
    
0 1 1 0 0 1
A= = .
1 0 0 1 1 0

4.2.6 We are given three vectors b1 , b2 , and b3 . The linear transformation L, mapping R2 to
R3 , is given by
L(x) = x1 b1 + x2 b2 + (x1 + x2 )b3 .
The problem is to find the matrix A representing
 L with respect to the bases [e1 , e2 ] and
B = [b1 , b2 , b3 ]. The matrix is A= a1 a2 , where ai = [L(ei )]B . Thus to find a1 we
first compute L(e1 ) = b1 + b3 ; the coordinate vector of b1 + b3 with respect to the basis
 B is (1, 0, 1)T . Similarly we find
L(e2 ) = b2 + b3 ; the coordinate vector with respect to
1 0
B is (0, 1, 1)T . So A = 0 1 .
1 1
Note: In the text, we are given concrete vectors b1 = (1, 1, 0)T , b2 = (1, 0, 1)T , and
b3 = (0, 1, 1)T . But, since the the description of L is in terms of b1 , b2 , and b3 , we need
not know what they look like. Had the description been in terms of, say, the standard
basis, this would not have been the case.

4.2.7 We are given vectors y1 = (1, 1, 1)T , y2 = (1, 1, 0)T , and y3 = (1, 0, 0)T ; these will be
used as an ordered basis for R3 . The identity operator on R3 is denoted I.

(a) Find the coordinates of I(ei ), i = 1, 2, 3, with respect to the basis [y1 , y2 , y3 ].
(b) Find a matrix A such that Ax is the coordinate vector of x with respect to Y =
[y1 , y2 , y3 ].

Solution:

(a) For each i = 1, 2, 3, we have I(ei ) = ei . We must express these standard basis vectors
with respect to the new basis. It is a simple matter (use Gaussian elimination
or simple observation) to show that [e1 ]Y = (0, 0, 1)T , [e2 ]Y = (0, 1, 1)T , and
[e3 ]Y = (1, 1, 0)T .

0 0 1
(b) From the preceding, it follows that A= 0 1 1 .
1 1 0
Note: This is simply change of basis couched in terms of linear transformations.

4.2.8a We have the same vectors yi as in problem 7. We are given a linear transformation
L : R3 R3 , where

L(c1 y1 + c2 y2 + c3 y3 ) = (c1 + c2 + c3 )y1 + (2c1 + c3 )y2 (2c2 + c3 )y3 .

We are to find a matrix A that represents L with respect to the ordered basis [y1 , y2 , y3 ].
We can do this in at least two ways.

(i) Not using the corollary: We are using the same basis throughout. For that reason,
using corollary 4.2.4 actually complicates things. In other words, only when we get
to part (b) of this problem does it make any difference what y1 , y2 , and y3 actually
look like. The first column of A is L(y1 ) = y1 + 2y2 ; the coordinates of this sum
with respect to Y are (1, 2, 0). The second column of A is L(y2 ) = y1 2y3 ; the
coordinates of this sum with respect to Y are (1, 0, 2). The third column of A is
L(y3 ) = y1 + y2 y3 ; the coordinates of this sum with respect to Y are (1, 1, 1).

2
 (ii) Using corollary 4.2.4: we have m = n and E = F = [y1 , y2 , y3 ]. The corollary leads
us to perform Gaussian elimination on the augmented matrix

1 1 1 3 1 1
[y1 y2 y3 |L(y1 )L(y2 )L(y3 )] = 1 1 0 3 1 2 .
1 0 0 1 1 1

The result of elimination is

1 0 0 1 1 1
[I|A] = 0 1 0 2 0 1 .
0 0 1 0 2 1

4.2.8(b) We are given a selection of vectors in R3 . It is very important to note from the
defintion of L, that the input basis and the output basis are both [y1 , y2 , y3 ] (that is the
case since the input is a linear combination of y1 , y2 , y3 , and so is the output). As a side
note, compare this defintion of L to the definition of L in problem 18 where the output
basis of L is standard basis.
Now, each given vector is however given to us in standard basis, so we must first put
that in the basis [y1 , y2 , y3 ], and then to use the matrix from part (a) to find the image
since the matrix in part (a) works from basis [y1 , y2 , y3 ] to basis [y1 , y2 , y3 ]. To do so, we
should find the transition matrix from the standard basis (since thats what the input is
given as in (i), (ii), and (iii) in the book) to the basis [y1 , y2 , y3 ]. That transition matrix
is Y 1 . Then the first of the given vectors is x = (7, 5, 2)T , so putting this vector in
the new basis, we either do [x]Y = Y 1 x = (2, 3, 2)T or we set up an augmented matrix
[y1 y2 y3 |x] whose solution is [x]Y = (2, 3, 2)T . Once the vector is in the desired basis as
required by the input of L, namely [x]Y , we can now apply the linear transformation:

[L(x)]Y = A[x]Y = (7, 6, 8)T .

A second method would be to generalize the multiplications above to:

1 1 1 0 0 1 7 7
[L(x)]Y = AY 1 x = 2 0 1 0 1 1 5 = 6 .
0 2 1 1 1 0 2 8

4.2.14(a) The linear operator L defined by

L(p(x)) = p0 (x) + p(0)

maps P3 into P2 . Find the matrix representation of L with respect to the ordered bases
[x2 , x, 1] and B = [2, 1 x]. Find the coordinates of L(p(x)) with respect to the given
basis for P2 if p(x) = x2 + 2x 3.

3
 Solution: We must first compute L(p(x)) for each basis polynomial p. Note that no
basis was specified for input or output, and so by default it will be the standard basis
for each (that is the input of l will have the standard basis [x2 , x, 1] and the output will
be in standard basis [x, 1]). Thus, we must then express each result with respect to the
given ordered basis for P2

1. L(x2 ) = 2x + 02 = 2x; [2x]B = (1, 2)T .

2. L(x) = 1 + 0 = 1; [1]B = ( 21 , 0)T .
3. L(1) = 0 + 1 = 1; [1]B = ( 12 , 0)T .
 
1 1/2 1/2
It follows that the matrix we want is A = . The coordinate vector
2 0 0
of p(x) = x2 + 2x 3 with respect to the given basis for P3 is z = (1, 2, 3)T . The
coordinate vector of L(p(x)) with respect to the basis B is therefore Az = (1/2, 2)T .
Thus L(x2 + 2x 3) = 12 (2) 2(1 x) = 2x 1, as expected.

4.2.18(a) We have bases E = [u1 , u2 , u3 ] and F = [b1 , b2 ], where u1 = (1, 0, 1)T , u2 =

(1, 2, 1)T , u3 = (1, 1, 1)T , b1 = (1, 1)T , and b2 = (2, 1)T . We are to find the matrix
A representing L(x) = (x3 , x1 )T with respect to E and F .
Solution: Now, in this problem the basis for the input of L is given to be E = [u1 , u2 , u3 ],
and the output is the standard basis for R2 by looking at the formula for L. We thus
first find the matrix for the transformation L, by applying L to the basis elements of the
input basis: L(ui ) for i = 1, 2, 3: we find that L(u1 ) = (1, 1)T , L(u2 ) = (1, 1)T , and
L(u3 ) = (1, 1)T . Again, this output is in standard basis for R2 , namely [e1 , e2 ]. And
so we need to find the change of basis formula from the standard basis to the new basis
F since the problem asked to find the matrix representing L with respect to the ordered
basis E and F .
Letting B be the matrix with columns b1 and b2 that form  the basis
 F , we find the
1 2
transition matrix from standard basis to F to be B 1 = . For each vector
1 1
1
L(ui ) that was in standard basis, we now calculate
 [L(ui )]B = B L(ui ) for i = 1, 2, 3;
1 3 1
the results are the columns of A = , which is the matrix representing L
0 2 0
with respect to the ordered basis E and F (input of L is in E and output of L is in F ).