# 7.3: Orthogonal Expansions

$$\newcommand{\vecs}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} }$$

$$\newcommand{\vecd}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash {#1}}}$$

$$\newcommand{\id}{\mathrm{id}}$$ $$\newcommand{\Span}{\mathrm{span}}$$

( \newcommand{\kernel}{\mathrm{null}\,}\) $$\newcommand{\range}{\mathrm{range}\,}$$

$$\newcommand{\RealPart}{\mathrm{Re}}$$ $$\newcommand{\ImaginaryPart}{\mathrm{Im}}$$

$$\newcommand{\Argument}{\mathrm{Arg}}$$ $$\newcommand{\norm}[1]{\| #1 \|}$$

$$\newcommand{\inner}[2]{\langle #1, #2 \rangle}$$

$$\newcommand{\Span}{\mathrm{span}}$$

$$\newcommand{\id}{\mathrm{id}}$$

$$\newcommand{\Span}{\mathrm{span}}$$

$$\newcommand{\kernel}{\mathrm{null}\,}$$

$$\newcommand{\range}{\mathrm{range}\,}$$

$$\newcommand{\RealPart}{\mathrm{Re}}$$

$$\newcommand{\ImaginaryPart}{\mathrm{Im}}$$

$$\newcommand{\Argument}{\mathrm{Arg}}$$

$$\newcommand{\norm}[1]{\| #1 \|}$$

$$\newcommand{\inner}[2]{\langle #1, #2 \rangle}$$

$$\newcommand{\Span}{\mathrm{span}}$$ $$\newcommand{\AA}{\unicode[.8,0]{x212B}}$$

$$\newcommand{\vectorA}[1]{\vec{#1}} % arrow$$

$$\newcommand{\vectorAt}[1]{\vec{\text{#1}}} % arrow$$

$$\newcommand{\vectorB}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} }$$

$$\newcommand{\vectorC}[1]{\textbf{#1}}$$

$$\newcommand{\vectorD}[1]{\overrightarrow{#1}}$$

$$\newcommand{\vectorDt}[1]{\overrightarrow{\text{#1}}}$$

$$\newcommand{\vectE}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash{\mathbf {#1}}}}$$

$$\newcommand{\vecs}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} }$$

$$\newcommand{\vecd}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash {#1}}}$$

##### Note

As stated in Section 7.2, the coefficients of \ref{eq:fourier} are defined as so:

$\label{ao} a_0=\dfrac{1}{L}\int_{-L}^{L}f(x)dx$

$\label{an} a_n=\dfrac{1}{L}\int_{-L}^{L}f(x)\cos{\left(\dfrac{n\pi x}{L} \right)}dx$

$\label{bn} b_n=\dfrac{1}{L}\int_{-L}^{L}f(x)\sin{\left(\dfrac{n\pi x}{L} \right)}dx$

The idea of expressing functions as a linear combination of the functions of a given basis set is more general than what we just saw. The sines and cosines are not the only functions we can use, although they are a particular good choice for periodic functions. There is a fundamental theorem in function theory that states that we can construct any function using a complete set of orthonormal functions.

The term orthonormal means that each function in the set is normalized, and that all functions of the set are mutually orthogonal. For a function in one dimension, the normalization condition is:

$\label{eq:fourier_normalization} \int_{-\infty }^{\infty }{\left | f (x) \right |}^2\; dx=1$

Two functions $$f(x)$$ and $$g(x)$$ are said to be orthogonal if:

$\label{eq:fourier_orthogonal} \int_{-\infty }^{\infty }{f (x) g^*(x) }\; dx=0$

The idea that you can construct a function with a linear combination of orthonormal functions is analogous to the idea of constructing a vector in three dimensions by combining the vectors $$\vec{v_1}=\{(1,0,0)\}, \vec{v_2}=\{(0,1,0)\},\vec{v_3}=\{(0,0,1)\},$$ which as we all know are mutually orthogonal, and have unit length.

The basis set we use to construct a Fourier series is

$\{1, \sin{(\frac{\pi}{L} x)}, \cos{(\frac{\pi}{L} x), \sin{(2\frac{\pi}{L} x)}, \cos{(2\frac{\pi}{L} x)}, \sin{(3\frac{\pi}{L} x)}}, \cos{(3\frac{\pi}{L} x)}...\} \nonumber$

We will prove that these functions are mutually orthogonal in the interval $$[0,2L]$$ (one period).

For example, let’s prove that $$\sin{(n\frac{\pi}{L} x)}$$ and $$1$$ are orthogonal:

$\int sin\left (\frac{n\pi x}{L} \right )dx=-\frac{L}{n\pi}cos\left (\frac{n\pi x}{L} \right ) \nonumber$

$\int_{0}^{2L} sin\left (\frac{n\pi x}{L} \right )dx=-\frac{L}{n\pi}cos\left (2n\pi \right )+\frac{L}{n\pi}cos(0)=\frac{L}{n\pi}\left ( 1-cos(2n\pi) \right )=0 \nonumber$

We can also prove that any $$\sin{(nx)}$$ is orthogonal to any $$\cos{(nx)}$$:

$\int sin\left (\frac{n\pi x}{L} \right ) \cos\left (\frac{n\pi x}{L} \right )dx=-\frac{L}{4n\pi}\cos\left (\frac{2n\pi x}{L} \right ) \nonumber$

$\int_{0}^{2L} \sin\left (\frac{n\pi x}{L} \right ) cos\left (\frac{n\pi x}{L} \right )dx=-\frac{L}{4n\pi}\cos\left (4n\pi\right )+\frac{L}{4n\pi}\cos (0)=0 \nonumber$

Following the same procedure, we can also prove that

$\int \sin\left (\frac{n\pi x}{L} \right ) \sin\left (\frac{m\pi x}{L} \right )dx=0\;n\neq m \nonumber$

$\int \cos\left (\frac{n\pi x}{L} \right ) \cos\left (\frac{m\pi x}{L} \right )dx=0\;n\neq m \nonumber$

The functions used in a Fourier series are mutually orthogonal. Are they normalized?

$\int_{0}^{2L} \sin^2\left (\frac{n\pi x}{L} \right )dx=L \nonumber$

$\int_{0}^{2L} \cos^2\left (\frac{n\pi x}{L} \right )dx=L \nonumber$

$\int_{0}^{2L} 1^2\;dx=2L \nonumber$

They are not! The functions $$1/2L, \frac{1}{L}\sin{(\frac{\pi}{L} x)}$$ and $$\frac{1}{L}\cos{(\frac{\pi}{L} x)}$$ are normalized, so we may argue that our orthonormal set should be:

$\{\frac{1}{2L},\frac{1}{L} \sin{(\frac{\pi}{L} x)},\frac{1}{L} \cos{(\frac{\pi}{L} x),\frac{1}{L} \sin{(2\frac{\pi}{L} x)}, \frac{1}{L}\cos{(2\frac{\pi}{L} x)}}, ...\} \nonumber$

and the series should be written as:

$\label{eq:fourier2} f(x)=c_0\frac{1}{2L}+\frac{1}{L}\sum_{n=1}^{\infty}c_n \cos\left ( \frac{n\pi x}{L} \right )+\frac{1}{L}\sum_{n=1}^{\infty}d_n \sin\left ( \frac{n\pi x}{L} \right )$

where we used the letters $$c$$ and $$d$$ to distinguish these coefficients from the ones defined in Equations \ref{ao}, \ref{an} and \ref{bn}.

However if we compare this expression to Equation \ref{eq:fourier}:

$\label{eq:fourier} f(x)=\frac{a_0}{2}+\sum_{n=1}^{\infty}a_n cos\left ( \frac{n\pi x}{L} \right )+\sum_{n=1}^{\infty}b_n sin\left ( \frac{n\pi x}{L} \right )$

we see that it is just a matter of how we define the coefficients. The coefficients in Equation \ref{eq:fourier} equal the coefficients in Equation \ref{eq:fourier2} divided by $$L$$. In other words, the coefficients in Equation \ref{eq:fourier} already contain the constant $$L$$ (look at Equations \ref{ao}, \ref{an} and \ref{bn}), so we can write the sines and cosines without writing the factor $$1/L$$ every single time.

In conclusion, the set

$\{1, \sin{\left(\frac{\pi}{L} x\right)}, \cos{\left(\frac{\pi}{L} x\right), \sin{\left(2\frac{\pi}{L} x\right)}, \cos{\left(2\frac{\pi}{L} x\right)}, \sin{\left(3\frac{\pi}{L} x\right)}}, \cos{\left(3\frac{\pi}{L} x\right)}...\} \nonumber$

is not strictly orthonormal the way it is written, but it is once we include the constant $$L$$ in the coefficients. Therefore, the cosines and sines form a complete set that allows us to express any other function using a linear combination of its members.

There are other orthonormal sets that are used in quantum mechanics to express a variety of functions. Just remember that we can construct any function using a complete set of orthonormal functions.

We can construct any function using a complete set of orthonormal functions.

This page titled 7.3: Orthogonal Expansions is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Marcia Levitus via source content that was edited to the style and standards of the LibreTexts platform; a detailed edit history is available upon request.