Comparison of vector algebra and geometric algebra

Content sourced from Wikipedia, licensed under CC BY-SA 3.0.

Geometric algebra vs vector algebra: A simple guide

Geometric algebra (GA) is an extension of ordinary vector algebra (VA). It adds new kinds of objects (like oriented areas and volumes) and a single, powerful product that unifies many familiar operations. GA works in any dimension and any signature, including 3D space and spacetime.

What GA adds to VA
- A single product, the geometric product, that combines two pieces of information: how much they point in the same direction and how their directions span a plane.
- New objects called multivectors. In 3D, these include scalars (grade 0), vectors (grade 1), bivectors (grade 2, representing oriented planes), and the pseudoscalar (grade 3, representing oriented volume). Bivectors and the pseudoscalar exist naturally in GA, rather than being added on later.
- A natural way to talk about areas and volumes. The magnitude of a bivector is an oriented area; the magnitude of a trivector (in 3D, built from three vectors) is an oriented volume.

The geometric product and its meaning
- For vectors A and B, the geometric product is AB = A · B + A ∧ B.
- A · B is the ordinary inner (dot) product.
- A ∧ B is the exterior (wedge) product, which encodes the oriented plane spanned by A and B.
- The wedge product generalizes the idea of the cross product. In 3D, the cross product u × v is the dual of the wedge product u ∧ v, using the unit pseudoscalar I (I = e1 e2 e3). This duality means GA treats the cross product as a natural partner to the wedge product, without needing a separate construction.

Key ideas in 3D GA
- The pseudoscalar I = e1 e2 e3 is a fundamental object in Euclidean 3-space. It provides a duality between bivectors (planes) and vectors (directions).
- The bivector subspace in 3D behaves like a three-dimensional vector space itself, with standard basis objects i = e2 e3, j = e1 e3, k = e1 e2.
- The commutator product A × B = (AB − BA)/2 captures the antisymmetric part of the geometric product. For the basic bivector basis, i × j = k, j × k = i, k × i = j, and the cross-relations are antisymmetric (e.g., j × i = −k). The commutator of equal bivectors is zero (i × i = j × j = k × k = 0).

Two-dimensional versus three-dimensional intuition
- In 2D, GA gives torque and curl as bivectors (a pseudoscalar in 2D), without introducing an external third dimension. This makes certain physical and geometric ideas more natural.
- In 3D, the cross product arises as a dual of a wedge product, tying together areas (bivectors), directions (vectors), and volumes (the pseudoscalar) in a single framework.

Projections, decompositions, and rotations
- GA makes projection and rejection (the part of a vector along a direction and the part perpendicular to that direction) look compact. If u is a nonzero vector, a vector v can be decomposed as
v = (u⁻¹)(u · v + u ∧ v),
where u⁻¹ is the inverse of u in the geometric sense.
- The projection part uses the dot product, and the rejection part uses the wedge product. Similar ideas extend to projecting onto planes and finding components orthogonal to a plane.
- These decompositions generalize to higher dimensions and connect naturally to how areas and volumes interact with vectors.

Wedge product, determinants, and linear systems
- The wedge product generalizes determinants and areas to higher dimensions. It offers a way to formulate solutions to linear systems without needing all the traditional determinant machinery.
- In GA, Cramer's rule-like results arise from wedge products. For example, solving a2x2 or a3x3 system can be expressed in terms of wedge products of the coefficient vectors, with the nonzero wedge factors playing the role of determinants.
- This approach extends beyond two or three variables and gives a unified geometric view of linear algebra.

Length, area, and volume
- In GA, the “length” of a vector is the usual Euclidean length.
- The “length” of a bivector is its oriented area.
- The “length” of a trivector (in 3D, built from three vectors) is its oriented volume.
- These measures come from the geometric product and its multivector components, tying together linear algebra and geometry in a single language.

Practical takeaways
- The geometric product AB encapsulates both projection (through A · B) and rotation-like or area-spanning behavior (through A ∧ B).
- Bivectors are the natural representatives of oriented planes; pseudoscalars give you oriented volumes. This unifies many quantities that in VA required separate constructs (like pseudovectors, curl, and torque).
- The cross product is not a standalone operation in GA; it is the dual of the wedge product in 3D, made simple and natural by the pseudoscalar I.
- The wedge product provides a powerful and compact way to reason about systems of equations and determinants, while also describing areas and volumes directly.
- GA provides a consistent framework in which many familiar operations—dot, cross, curl, projections, rotations, and solving linear systems—are different manifestations of the same underlying geometric product.

Why use geometric algebra
- It unifies many disparate tools into one framework, reducing the need for ad hoc constructions.
- It gives a clear geometric interpretation to algebraic operations, linking vectors, areas, and volumes in a single language.
- It naturally handles higher dimensions and different signatures, making it a flexible tool for physics, engineering, and computer science.

In short, geometric algebra extends vector algebra by introducing new objects (bivectors, trivectors) and a single, unifying product (the geometric product) that encodes both length and orientation. It provides a natural, compact way to handle directions, planes, and volumes, while preserving and generalizing the familiar operations you already know from vector algebra.