Orthogonality of Trigonometric Functions
You'll learn: How sine and cosine functions behave like perpendicular vectors, enabling us to decompose complex periodic functions into simple building blocks.
The Core Idea
Just as perpendicular vectors in geometry have a dot product of zero, certain pairs of trigonometric functions are "perpendicular" (orthogonal) over an interval. This means their inner product equals zero.
For functions, the inner product is defined as an integral:
⟨f, g⟩ = ∫[a to b] f(x)g(x) dx
Two functions are orthogonal on interval [a, b] if ⟨f, g⟩ = 0.
Why This Matters
On the interval [-π, π] (or any interval of length 2π), the sine and cosine functions form an orthogonal family:
- ∫[-π to π] sin(nx) sin(mx) dx = 0 when n ≠ m
- ∫[-π to π] cos(nx) cos(mx) dx = 0 when n ≠ m
- ∫[-π to π] sin(nx) cos(mx) dx = 0 for all integers n, m
When n = m, these integrals yield π (not zero), which measures the "length" or norm of each function.
The Analogy
Think of building with LEGO bricks aligned along perpendicular axes. Each sine and cosine function is like a direction in an infinite-dimensional space. Because they're orthogonal, you can build any periodic function by stacking these components independently—just like decomposing a 3D vector into x, y, and z components.
This orthogonality is what makes Fourier series work: you can determine each coefficient independently by taking inner products with the corresponding basis function.
Key Takeaway: Sine and cosine functions are orthogonal under integration over appropriate intervals, allowing us to use inner products to extract individual Fourier coefficients from periodic functions.