Recall from Yesterday: Coordinate

Systems
• Observation 1: Given a basis for a finite-dimensional vector space
V, we recall that a vector can be expressed in one and only one way
as a linear combination of the basis vectors. This motivates the
following definitions:
• Definition: An ordered basis for a finite-dimensional space V is a
finite sequence of vectors which is linearly independent and spans
V. In other words, an ordered basis is a basis with the vectors taken
in a specified fixed order.
• Given an ordered basis B = {u1,u2,….,un} we can express any
vector uniquely in the form u = x1 u1 + x2u2 + ….. + xn un. The
scalars xi are called the coordinates of u relative to the (ordered)
basis B.
 Coordinate Mapping - 1
• Observation 2: Given a fixed ordered basis B for a finite-dimensional vector
space V, we can therefore find an n-tuple in Fn (here F denotes the field of
scalars; for us it would usually be either R or C) corresponding to any vector in V
as follows :
• u  ( x1, x2,…..,xn), where the xi are the coordinates of u relative to B, i.e. the n-
tuple we take is precisely the coordinates of the vector u taken in order.
• Rather than the n-tuple ( x1, x2,…..,xn), for convenience we express
this in a column format, i.e. we prefer to take the column vector:
 x1
| x2 |
| : |
 xn 
• This vector is called the coordinate vector of u (relative to B) or the B-
coordinate vector of u and is written [u]B . In some books it is called the
coordinate matrix .
 Coordinate Mapping - 2
• The mapping or correspondence u  [u]B is called the coordinate mapping
determined by B. It has the following properties:
• It is a one-to-one correspondence, i.e. each vector has a unique
corresponding n-tuple (sequence), and each n-tuple (sequence) has a unique
corresponding vector.
• The sum of two vectors corresponds to the sum of the two n-tuples
• The scalar multiple of a vector corresponds to the scalar multiple of the n-
tuple.
• The brief discussion above (details left as an exercise) indicates that the
coordinate mapping is actually an isomorphism from an n-dimensional vector
space V over the field F to Fn.
• NB: We have already seen by a general proof that V and Fn are isomorphic. The
above approach gives us a way to define a definite isomorphism. Note that we get
a different isomorphism for each choice of an ordered basis for V.
 The Matrix of a Linear Transformation - 1
• We proceed as follows. Suppose V and W are finite-
dimensional vector spaces over the field F, and T: V W
is a linear transformation . We will associate a matrix with
this linear transformation.
• Suppose dim V = n and dim W = m.
• We take a fixed ordered basis B = {v1,v2,…,vn} for V, and
a fixed ordered basis C = {w1,w2,…,wm} for W.
• Since Tvi belongs to W, we can express it uniquely as a
linear combination of the wj:
Tvi = A1iw1 + A2iw2 + …. + Amiwm
We now form the mn matrix A with these coefficients as
columns.
• The matrix A is called the matrix of T with respect to the
bases B and C. We also use the notation: [T] BC
 The Matrix of a Linear Transformation – Example

1. To find the matrix of the linear transformation T: R3  R2 given by

T(x,y,z) = (x + y + z, x + 2y + 3z), with respect to the standard
bases for R3 and R2 respectively.
T(1,0,0) = (1,1) = 1e1 + 1e2
T(0,1,0) = (1,2) = 1e1 + 2e2
T(0,0,1) = (1,3) = 1e1 + 3e2

Hence the matrix of T is A = 1 1 1

 1 2 3
For any vector v = (x,y,z) in R3 expressed in column form we have:
Av = A x  = x + y + z = Tv
| y |  x + 2y + 3z
z