which pr

10.14. GRAM-SCHMIDT ORTHOGONALIZATION PROCESS

Theorem. Every finite dimensional inner product space has an orthonormal

basis.

M.D.U. 2016, 13, 09, 06; K.U. 2014, 12]

Proof. Let V(F) be an inner product space of dimension n
Let u, u, be basis of V.
S =
U..., a
C lo28)
Since every orthonormal set is linearly independent and dim V=n,it is enough to construct
an orthogonal basis ofV.
Now, S being a basis is linearly independent

u 0 for i =
1, 2,.., n

Let <ug, U >U1

1 UDefine v2 =
ug

Now U = 0

u , U U1 scalar multiple of vi
Il,
or
u2 Scalar multiple of u
 LINEAR ALGEBRA
10.38
is linearly dependent
l u , uzl
contradiction, as
lu, u being a subset of a linearly independent set S.
Sis
which is a
independent.
inearly
Thus Uo*0

Then ua-U2, U1u, >=<u

U2 U>= <u2- <Ug,
->2 , 1,u
u>
>

=
<ug, u >-
<U, [: <up U1>
ll
=
| u, 11
<Ug, U >=<u2, -<u2, u>
= 0

U2 is orthogonal to u, and so
lu,, U2} is a orthogonal set.
Define v, = <ug, U1 2 U1 u3, U2U2
i=1
Again ug +0 as u, = 0
{u1, Ug, ul is linearly dependent.
But lu, u2, ug} being a subset of
linearly independent set S is linearly independent.
Then U3, U >= <uz

=<uz, U >-u3: O1?5U1s U1

<U3, Ua < U2, U1>
<ug 1 >-<uz, U>-0 =0D
In a similar way, we note that
:<vp U,> =
|| U, and <
<
Ug, U,> = 0 =
U2 U,> =0
Thus <U, <Uz» V>
V,> =0, where i *j and
i,j 1, 2, 3. =

In this way, we can construct an

orthogonal set (u,, U2-t., U, of n
vectors.
Take w U\
Then w, Wp..,
Uw,} is an orthonormal
set of vectors
Since an orthonormal set of in V.
vectors in an
so {w,,
w,..., wl is linearly inner product
independent. space is a linearly
orthonormal basis ofV. Also dim V
=n, thus independent set, (w1 W2 w,} forms an
INNER PRODUCTSPACES
10.39

Example22.

Obtain an orthonormal basis uvith respect to standard inner product for the
hspace of r 'generated by (1, 0, 1), (1, 0, - 1) and (0, 3, 4). K.U. 2017, 12]
Solution. Let u = , Xg g), v = 12 ) e R". Then by definition ofinner product

< u, U> =
*y1 +
2y2 + *3Y3 .(1)

Let u =
(1, 0, 1), u2 =
(1, 0, 1) and ug
-
=
(0, 3, 4)
Using (1), we get

U = <u, u> =1+0+ 1 =2 ..(2)

ll u2 IP= < uz, luz > = 1+0+1 = 2 ..(3)

| 43 l1 = <ug, ug> = 0+9+ 16 =25 4)

U , U> =
1+0-1=0 ..(5)
...6)
Ug, U> = 0 +0+4 4
...(7)
Ug, Ug> =0+0-4 =-4
ug, U U
Define U1 u1, U2 =u2

and
Ug, U1>U1KUs, V2 >U2
U3 u3

2
<u, l=u2- 0 2 u2 By (2) and (5))

i.e. U2
<ug,1> U1<Ug,u2 > u2
Also, U3 U 3

_-4)ug2 By (2), (3), (6) and (7))

UgUg - 2

ug-2u + 2u,2

0 , 3, 4)-21,0, 1) + 201,0,-1) =(0-2+2,3-0+0, 4-2-2)

(0, 3, 0)
.e. Ug

V =U1 U2 =
ug, Vg =
(0, 3, 0)
Thus

ll D l= 0 +9+0 =3
10.40 LINEAR ALGEBRA
U2 U3
Put
T W2
T W

The required orthonormal basis is

Uaing Gran Sehmuelt Onhogaraligaken
Corshuet a n Onttwnormal
mocess,
basis of V, (R) urith standass innar

rduck dapinad Sn it, given a

basis
uCu,1), 4(l,2,1),
u ( 2,3).
So Det V, = u (h'}, 1)

V2

C1-2,1), ti,1,1)2
le to)+(-2)I)+C1)1)
-2+!

weha 23), 1
<, 3

Al,y,Va
Al ua,1), ti,D
E - 4 +3 =o
 Ustt cadeutated value, un

V -
3

te.V '2,3)-C+,)
iie. V ( 4 , °,1)
t D , 2,/), (ho,1)
ettwnsmal bas's ls
T eguired

()»(thah}