Motivation for Orthonormal Bases
You'll learn why orthonormal bases are the "gold standard" in linear algebra, making calculations simpler and more reliable.
The Computational Advantage
Imagine you're working with coordinates in a badly skewed coordinate system where axes point in awkward directions and have different scales. Every calculation—finding lengths, angles, projections—becomes messy. Now imagine switching to a nice perpendicular grid where each axis has length 1. Suddenly everything becomes cleaner.
That's the power of orthonormal bases.
Why Orthonormal Bases Simplify Computations
When you have an orthonormal basis {u₁, u₂, ..., uₙ}, computing coordinates is trivial. For any vector v, its coordinates are simply:
Coordinate cᵢ = ⟨v, uᵢ⟩
No matrix inversions, no systems of equations—just inner products. Compare this to a general basis, where you'd need to solve a system or compute a matrix inverse to find coordinates.
Computing norms is equally simple: ||v||² = c₁² + c₂² + ... + cₙ² (just the sum of squared coordinates). With a non-orthonormal basis, you'd need to account for cross-terms from the Gram matrix.
Numerical Stability
In practical applications, orthonormal bases provide crucial numerical stability. When computers perform arithmetic with finite precision, small errors can compound. Operations involving orthonormal bases are inherently well-conditioned because:
- All basis vectors have the same length (no huge or tiny numbers)
- Vectors are perpendicular (no near-linear dependencies that amplify errors)
- Transformations preserve lengths and angles
This makes orthonormal bases essential in numerical linear algebra, computer graphics, signal processing, and scientific computing.
Key Takeaway: Orthonormal bases transform complex linear algebra operations into simple, numerically stable calculations—they're perpendicular, unit-length, and perfect for both theoretical elegance and practical computation.