Professional Documents
Culture Documents
Radu MICULESCU
octomber 2019
g (x, y ) = g (y , x ),
for every x, y 2 V .
B = fe1 , ..., en g
a basis of V ,
n
x= ∑ xk ek 2 V ,
k =1
n
y= ∑ yl el 2 V
l =1
2 3 1
A=( 3 1 1 ).
1 1 2
It follows that
0
rank (A) = rank (A )
and this value de…nes the rank of the bilinear form g .
2 3 1
( 3 1 1 ).
1 1 2
where
f1 = (1, 1, 1), f2 = (1, 1, 0) and f3 = (1, 0, 0),
and the quadratic form associated to g .
0
The matrix of change from the canonical basis to B is
1 1 1
Ω = ( 1 1 0 ).
1 0 0
0
Hence the matrix of g with respect to B is
7 7 4
t
ΩAΩ = ( 7 9 5 ).
4 5 2
Remarks.
1. ker(g ) is a subspace of V .
2.
rank (g ) + dim ker(g ) = dim V .
3. g is nondegenerate if and only if ker(g ) = f0V g.
h(x ) = g (x, x ),
for every x 2 V .
Remarks.
1. The symmetric bilinear form g which uniquely de…nes the quadratic
form h is called the polar form of h.
2. If the quadratic form h is known, then the associated polar form g is
given by the following expression:
1
g (x, y ) = [h (x + y ) h (x ) h(y )],
2
for every x, y 2 V .
R. Miculescu (Institute) Bilinear and quadratic forms octomber 2019 13 / 42
Example
for every x = (x1 , ..., xn ), y = (y1 , ..., yn ) 2 Rn de…nes the quadratic form
h : Rn ! R given by
i.e.
n
h(x1 , ..., xn ) = ∑ xk2 ,
k =1
g (x, y ) = 0.
g (x, y ) = 0,
for every x, y 2 U, x 6= y .
Let
n n
∑ ∑ akl xk xl
k =1 l =1
i ,j =1
0
with aii 6= 0.
Hence, without loss of generality, we can assume that
0
a11 6= 0.
k =2 i ,j =2
a11 i ,j =2
By using Gauss method …nd the diagonal form of the quadratic form
h : R3 ! R whose analytic expression with respect to the canonical basis
is given by
h(x ) = x1 x2 + x1 x3 + x2 x3 .
0 0
where xk s are the coordinates of x in the basis B and δ0 = 1.
R. Miculescu (Institute) Bilinear and quadratic forms octomber 2019 25 / 42
Carl Gustav Jacob Jacobi
1804-1851
By using Jacobi method …nd the diagonal form of the quadratic form
h : R3 ! R whose analytic expression with respect to the canonical basis
is given by
1 8 2
( 8 8 2 ).
2 2 1
By using the eigenvalues method …nd the diagonal form of the quadratic
form h : R3 ! R whose analytic expression with respect to the canonical
basis is given by
3 2 4
( 2 6 2 ).
4 2 3
3 λ 2 4
2 6 λ 2 = 0,
4 2 3 λ
i.e.
λ1 = λ2 = 7 and λ3 = 2.
Therefore we get the following canonical form of h:
0 0 0
7 ( x1 ) 2 + 7 ( x2 ) 2 2 ( x3 ) 2 .
a canonical form of h.
Then we shall use the following notations:
not
#(fk 2 f1, ..., ng j ak = 0g) = z
not
#(fk 2 f1, ..., ng j ak > 0g) = p
not
#(fk 2 f1, ..., ng j ak < 0g) = q.
p is called the positive index of inertia, q is called the negative index of
inertia and the pair (p, q ) is called the signature of h.
z is called the nullity of h and it is n rank (h).
R. Miculescu (Institute) Bilinear and quadratic forms octomber 2019 32 / 42
Sylvester’s law
5 2 2
( 2 6 0 ),
2 0 4
so
5 2 2
5 2
δ1 = j5j = 5, δ2 = = 26 and δ3 = 2 6 0 = 80.
2 6
2 0 4