Change of Basis

Prof. Dr. Shailendra Bandewar

In many applications it is observed that the
problem can easily solved by changing the
coordinate system. We know that basis provides
us the coordinate system for a vector space,
via the notion of coordinate vectors. Choosing
the right basis will often simplify a particular
problem.
 Coordinate Relative to Standard Basis
An ordered basis for a vector space V is an
Ordered set of linearly independeant set of
vectors which spans V .
Example 1 : Let e1  1, 0  , e2   0, 1 .
T T

Then B1  e1 , e2  and B2  e2 , e1 are two

different ordered basis for  2 .
Definition : (Coordinate of a vector realtive to
an ordered basis )
Let B  v1 , v2 , ...., vn  be a ordered basis of a
vector space V . Let v  V
 v  c1v1  c2 v2  ...  cn vn , where ci  .
then  c1 , c2 , ... , cn  is called the coordiante
T

vector of v relative to the basis B. It is denoted

by  v B .
Example 2 : Let e1  1, 0  , e2   0, 1 .
T T

Then B1  e1 , e2  and B2  e2 , e1 are two

different ordered basis for  . if v   2, 3 ,
2 T

then find  v B and  v B .

1 2

2 3 
  v B    and  v B    .
1
 
3 2
 2
Example 3 : Let e1  1, 0  , e2   0, 1 ,
T T

v1  1, 1 and v2  (0, 1) .

T T

If B1  e1 , e2  and B2  v1 , v2  if v   2, 3 ,

T

then find  v B and  v B .

1 2

2  2
  v B    and  v B    .
1
 
3 2
1 
as (2, 3)  c1 (1, 1)  c2 (0, 1)
 c1  2, c2  1
 e2

1 e1

Graph relative to the basis Graph relative to the basis

B1  (1, 0), (0, 1) B1  (1, 1), (0, 1)
Example : Consider Bases B  b1 , b2  and C  c1 , c2 
for a vector space V . Such that
b1  4c1  c2 and b2  2c1  3c2 and suppose
x  2b1  b2
2 
i.e  x B    . find  x C
 1
4  2 
 b1 C    and b2 C   
1  3 
  x C  2 b1 C  b2 C  b1 C b2 C   x B
 4 2   2  10 
     
1 3   1  1
Definition :
Let B  b1 , b2 , ... , bn  and C  c1 , c2 , ...., cn  be
ordered basis of a vector space V . Then there exists
a unique n  n matrix P such that
C B

 x C  CP B  x B
The columns of P are C  coordinate vectors of
C B

the vectors in the basis B. That is

P  b1 C b2 C ..... bn C 
C B
Theorem :
Let B  b1 , b2 , ... , bn  and C  c1 , c2 , ...., cn  be
ordered basis of a vector space V . and let P be
C B

the change of basis matrix from B to C. Then

1. P
C B
 x B   x C for all x  V .
2. P  x B is a unique matrix P with the property
C B

that P  x B   x C for all x  V .

C B
1
3. P is invertible and  P   P .
C B  C B  B C
Remark : Change of basis is a transformation
form  n to itself that simply switches from one
coordinate system to another.

x
V
 C  B
Multiplication by
P
C B

 xC  x B
 n n
Multiplication by
1
 P   P
 C B  B C
Example : Find the change of matrix P and
C B

P and B  1, x, x 2
 and C  1  x, x  x , 1  x 
2 2
B C

of P2 . Then Coordinate vector of p( x)  1  2 x  x 2

with respect to C.

 Changing to a standard basis is easy, so we first

find P . observe that coordinate vectors to C in
B C

terms of B are
 1  0 1 
     
1  x B  1  ,  x  x   1  , 1  x   0 
2 2

0  1  1 

1 0 1   1 1 1
1
 
P  1 1 0  and P  P   1
  1 1 1  
B C C B  B C  2
0 1 1   1 1 1 
 1 1 1 1   2 
1     
 p( x)c  CPB  p( x)B   1 1 1   2   0 
2
 1 1 1   1  1
Example : Let M 22 , Let B be the the standard basis
of M 22 and let C be the basis
C   A, B, C , D
 1 0 1 1  1 1  1 1 
   0 0  , 1 0  ,
, 1 1 , 
 0 0       
Find the change of matrix P and
C B

1 2 
verify that  x C  P  x B for x    .
C B
3 4 
 1 0 
 E11   
0 0
1 0  1 1  1 1  1 1
 a  b  c  d 
0 0 0 0 1 0  1 1
1 
0
  E11 C   
0
 
0
similarly
 0 1 
E12   
 0 0 
1 0  1 1  1 1  1 1
 a  b  c  d 
 0 0   0 0  1 0  1 1
 a  b  c  d  0, b  c  d  1, c  d  0, d  0
 a  1, b  1, c  0, d  0
 1
1 
  E21 C  
0 
 
0 
 0 0
E11   
1 0 
1 0  1 1  1 1  1 1
 a  b  c  d 
 0 0   0 0  1 0  1 1
 a  b  c  d  0, b  c  d  0, c  d  1, d  0
 a  0, b  1, c  1, d  0
0 
 1
  E21 C  
1 
 
0 
 0 0
E22   
0 1 
1 0  1 1  1 1  1 1
 a  b  c  d 
0 0 0 0 1 0 1 1
 a  b  c  d  0, b  c  d  0, c  d  0, d 1
 a  0, b  0, c  1, d  1
0 
0 
  E21 C  
 1
 
1 
 1 1 0 0 
0 1 1 0 
 P  
C B 0 0 1 1
 
0 0 0 1 
1 2 
If x   
 3 4 
1 
2
  x B   
3 
 
4
 1 1 0 0  1 
0 1 1 0   2 
 x c  P  x B    
C B 0 0 1 1 3 
  
0 0 0 1   4
 1
 1
 
 1
 
4 
Method 2 :
If E is another basis of M 22

P 
1
then P  P P  P .
C B C E EB E C EB
  1   2  
4. Let B and C be the bases for  . If C     ,   
2

  2  3  
 1 1
and change bases matrix from B to C is P   
C B
 1 2 
Find B.
 1   1  1   2 
 We have b1 C    and b2 C    If C     , 3  
 1 2   2  
1   2   1
 b1  1    1     
 2  3   1
1   2  3 
and b2  1    2     
2 3   4 
  1 3  
 B    ,   
  1  4  
Exercise
(a ) Find the coordinate vectors  x B and  x C
with respect to the bases B and C rspectively.
(b) Find the cahnage of matrix P .
B C

3  0 0 1  
    0 , 0 
1. x  1  , B   1  ,    
5 0 1   0 
 
 1  0 1  
  1  , 0  .
C   1  ,    
 0 1  1  
 
2. p ( x)  4  2 x  x 2 , B   x, 1  x 2 , x  x 2 
C  1, 1  x, x  .
2

1 1  1 0   0 1  1 1  1 0 
3. A    , B    ,  ,  ,  
1 1  0 1  1 0   0 0 1 0 
 1 1 1 1  1 0   0 1 
C    ,  ,  ,  .
 0 1 1 0  1 1  1 1 
4. Let B and C be the bases of P2 . If
B   x, 1  x, 1  x  x 2  and change of basis matrix
from B to C is
 1 0 0
P   0 2 1 
C B
 1 1 1 
Find C.
Example : Let P1 (t )  1  t 2 , p2 (t )  2  t  3t 3 ,
p3 (t )  1  2t  4t 2
a. Use Co  ordinate vectors to show that these
polynomials form a basis of P2 .
b. Consider the basis B   p1 , p2 , p3  for P2 . Find q
 3 
in P2 . Given that  q B   1  .
 2 
 1   3 3  8
Example : Let b1  0  , b2   4  , b3   6  , and x   2 
0   0   3   3 
a. Show that the set B  b1 , b2 , b3  is a basis of 3 .
b.Find the change of coordinates matrix from B to
the standard basis.
c.Wrte the equation that relates x in R 3 to  x B .
d . Find  x B , for the given x above.