Let
be a finite dimensional inner product space.
Suppose
is a linearly independent subset of
Then the Gram-Schmidt orthogonalisation
process uses the vectors
to construct new
vectors
such that
for
and
for
This process
proceeds with the following idea.
Suppose we are given two vectors
and
in a plane. If we
want to get vectors
and
such that
is a unit
vector in the direction of
and
is a unit vector
perpendicular to
then they can be obtained in the following
way:
Take the first vector
Let
be the angle between the
vectors
and
Then
Defined
Then
is a vector perpendicular
to the unit vector
, as we have removed the component of
from
.
So, the vectors that we are interested in are
and
This idea is used to give the Gram-Schmidt Orthogonalisation process which we now describe.
For
we have
Since
and
Hence, the result holds for
Let the result hold for all
That is, suppose we are given any set of
linearly independent vectors
of
Then by the inductive assumption,
there exists a set
of vectors satisfying the following:
Now, let us assume that we are given a set of
linearly independent vectors
of
Then by the inductive assumption,
we already have vectors
satisfying
On the contrary, assume that
Then
there exist scalars
such that
So, by (5.2.2)
Thus, by the third induction assumption,
This gives a contradiction to the given assumption that the set of vectors
So,
. Define
.
Then
. Also, it can be easily verified that
for
.
Hence, by the principle of mathematical induction, the proof of the theorem is complete.
height6pt width 6pt depth 0pt
We illustrate the Gram-Schmidt process by the following example.
Hence,
![]() |
![]() |
![]() |
|
![]() |
![]() |
We claim that in this case,
Since, we have chosen the smallest
satisfying
for
As
So, by definition of
Therefore, in this case, we can continue with the Gram-Schmidt process
by replacing
by
Let
is an
Also, observe that the conditions
and
for
implies that
![]() |
![]() |
![]() |
|
![]() |
![]() |
Perhaps the readers must have noticed that the inverse of
is its transpose. Such matrices are called
orthogonal matrices and they have a special role to play.
It is worthwhile to solve the following exercises.
where
Prove that
In case,
is non-singular, the diagonal entries of
can
be chosen to be positive. Also, in this case, the decomposition is
unique.
Let the columns of
be
The Gram-Schmidt
orthogonalisation process applied to the vectors
gives the vectors
satisfying
By using (5.2.5), we get
![]() |
![]() |
![]() |
|
![]() |
![]() |
||
![]() |
![]() |
The proof doesn't guarantee that for
is positive. But this can be achieved by replacing the vector
by
whenever
is negative.
Uniqueness: suppose
then
Observe the following properties of
upper triangular matrices.
Suppose we have matrix
of dimension
with
Then by Remark
5.2.3.2, the application of the Gram-Schmidt
orthogonalisation process yields a set
of orthonormal vectors of
In this case, for each
we have
Hence, proceeding on the lines of the above theorem, we have the following result.
We now compute
If
we denote
then by the Gram-Schmidt process,
and
The readers are advised to check that
Let
Define
Let
Then
Hence,
So, we again take
So,
The readers are advised to check the following:
A K Lal 2007-09-12