Vector space of functions

We study the collection of functions R2π2 (square-integrable functions with period 2π) as a vector space. We define the L2 norm and L2 inner product on this vector space so that we can investigate the ``geometric'' structure of the space of functions. 


Let us show that R2π2, the set of square-integrable functions with period 2π, is a vector space over C. First, we need to define addition and scalar multiplication. Let f,gR2π2. We define f+gR2π2 by

(eq:add)(f+g)(x)=f(x)+g(x), xR.
Note that the ``+'' on the left-hand side is defined between the two functions f and g, whereas the ``+'' on the right-hand side is the addition between two complex numbers f(x) and g(x). Next, we define scalar multiplication. Let αC and fR2π2. We define αf by
(eq:scale)(αf)(x)=αf(x).
Note that the product on the left-hand side is between the scalar α and the function f, whereas the product on the right-hand side is between the two complex numbers α and f(x).

Lemma 

R2π2 is a vector space with vector addition (eq:add) and scalar multiplication (eq:scale).
Proof. We show that R2π2 satisfies all the axioms of the vector space. In the following, f,g,hR2π2 and α,βC.

1. R2π2 is closed under vector addition.

Suppose f,gR2π2. We show f+gR2π2.

By assumption, ππ|f(x)|2dx<+ and ππ|g(x)|2dx<+. Thus,

ππ|f(x)+g(x)|2dxππ(|f(x)|+|g(x)|)2dx (the triangle inequality)=ππ|f(x)|2dx+ππ|g(x)|2dx+2ππ|f(x)||g(x)|dxππ|f(x)|2dx+ππ|g(x)|2dx+2ππ|f(x)|2+|g(x)|22dx=2(ππ|f(x)|2dx+ππ|g(x)|2dx)<+ (by assumption).

2. Addition is commutative.

For any f,gR2π2,

(f+g)(x)=f(x)+g(x)=g(x)+f(x)=(g+f)(x).

3. Addition is associative.

For any f,g,hR2π2,
((f+g)+h))(x)=(f+g)(x)+h(x)=(f(x)+g(x))+h(x)=f(x)+(g(x)+h(x))=f(x)+(g+h)(x)=(f+(g+h))(x).

4. The existence of the zero element (additive identity).

Let 0(x)=0 be the identically zero function. Clearly, 0R2π2, and for any fR2π2,
(0+f)(x)=0(x)+f(x)=f(x).

5. Additive inverses exist.

For any fR2π2, define f by (f)(x)=f(x). Clearly, fR2π2, and 
(f+(f))(x)=f(x)+(f(x))=0=0(x).

6. R2π2 is closed under scalar multiplication.

For each αC and fR2π2,
ππ|αf(x)|2dx=|α|2f2<+.

7. Multiplication by the sum of scalars: (c+d)f=cf+df.

For each fR2π2 and α,βC,
((α+β)f)(x)=(α+β)f(x)=αf(x)+βf(x)=(αf)(x)+(βf)(x)=(αf+βf)(x).
Thus, (α+β)f=αf+βf.

8. Scalar multiplication of the sum of vectors: α(f+g)=αf+αg.

For all f,gR2π2, and αC,
(α(f+g))(x)=α(f(x)+g(x))=αf(x)+αg(x)=(αf+αg)(x).

9. The existence of the multiplicative identity scalar.

For any fR2π2, 1C,
(1f)(x)=1f(x)=f(x).

10. Scalar multiplication is associative: (αβ)f=α(βf).

For any α,βC, and fR2π2
((αβ)f)(x)=(αβ)f(x)=α(βf(x))=α(βf)(x)=(α(βf))(x).

Thus, R2π2 is a vector space over C. ■

Definition (L2-norm, mean-square norm)

Let f be a function on (π,π) that is square-integrable, i.e.,
ππ|f(x)|2dx<+.
We define the mean-square norm or L2 norm f by
f=ππ|f(x)|2dx.

Remark. The L in L2 stands for ``Lebesgue.'' suggesting that we should use the Lebesgue integral rather than the Riemann integral. But the term ``L2'' is so widespread that we use it, although we only use the Riemann integral. □

See also: Norm (Wikipedia)

Definition (Normed space)

A vector space on which a norm is defined is called a normed vector space or, simply, normed space

Definition (L2 inner product)

Let f,g be complex-valued, square-integrable functions on (π,π). We define the L2 inner product between f and g as
(f,g)=ππf(x)g(x)dx
where g(x) is the complex conjugate of g(x)

We need to verify that the integral on the right-hand side of the above equation does exist. In fact,
|ππf(x)g(x)dx|ππ|f(x)g(x)|dx=ππ|f(x)||g(x)|dxππ|f(x)|2+|g(x)|22dx=12ππ|f(x)|2dx+12ππ|g(x)|2dx<+

Remark. The L2 inner product corresponds to the scalar (dot) product in a vector space. □

In general, an inner product is defined as follows.

Definition (Inner product (general))

Let V be a vector space over the field K. An inner product (,):V×VK is a map with the following properties for all vectors x,y,zV and all scalars α,βK:
  1. (Conjugate symmetry) (x,y)=(y,x).
  2. (Linearity in the first argument) (αx+βy,z)=α(x,z)+β(y,z)
  3. (Positive definiteness) If x0, (x,x)>0.

Example. Consider the vector space Rn. For x=(x1,x2,,xn),y=(y1,y2,,yn)Rn, the scalar product is defined as
(x,y)=i=1nxiyi.
For each aR, its ``conjugate'' is the same a: a=a. Thus, 
(y,x)=(y,x)=i=1nyixi=i=1nxiyi=(x,y).
For a,bR and x,y,zRn,
(ax+by,z)=i=1n(axi+byi)zi=ai=1nxizi+bi=1nyizi=a(x,z)+b(y,z).
If xRn is non-zero, at least one of its components, x1,x2,,xn, is non-zero. Thus,
(x,x)=i=1nxi2>0.
Thus, the scalar product in Rn is an inner product. □

Exercise. Show that L2 inner product satisfies the conditions of an inner product. □

Definition (Inner product space)

A vector space on which an inner product is defined is called an inner product space.

Example. R2π2 with the L2 inner product is an inner product space. □

Lemma (Cauchy-Schwarz inequality)

For f,gR2π2, we have
|(f,g)|fg.
Proof. Exercise. ■

Exercise. Using the Cauchy-Schwarz inequality, show the triangle inequality:
f+gf+g.

Theorem 

Let f,f1,f2,g,g1,g2 be square-integrable and αC be an arbitrary constant. The following hold:
f2=(f,f),(f,g)=(g,f),(f1+f2,g)=(f1,g)+(f2,g),(f,g1+g2)=(f,g1)+(f,g2),(αf,g)=α(f,g),(f,αg)=α¯(f,g).
Proof. Exercise. ■

Using the inner product, the Fourier coefficient of f can be written as
ck=12π(f,eikx)   (kZ).
The orthogonality of {einx} can be expressed as
(einx,eimx)=2πδn,m
where δn,m is Kronecker's delta.

Remark. If the function f has a period of 2π, the range of integration in
f2=ππ|f(x)|2dx
can be anything, i.e., ab|f(x)|2dx as long as ba=2π. The same argument applies to the inner product.

Comments

Popular posts from this blog

Birth process

Branching processes: Mean and variance

Informal introduction to formal logic