Overview

A unit vector is any vector whose length (magnitude) equals one. Unit vectors indicate direction without scaling: they carry orientation information but no magnitude. In formulas and prose, unit vectors are often denoted with a circumflex (hat) over the symbol, for example v̂. The term is fundamental in linear algebra, geometry and applied fields such as physics and computer graphics. Circumflex notation and boldface conventions are both used to mark unit vectors in printed material.

Definition and calculation

Given a nonzero vector v, its unit vector in the same direction is obtained by dividing v by its length (Euclidean norm). If ‖v‖ denotes the length, then the normalized vector is written v̂ = v / ‖v‖. This operation is called normalization. A simple example in two dimensions: the vector (3, 4) has length 5, so its unit vector is (3/5, 4/5). The zero vector cannot be normalized because its length is zero.

Key properties:

  • The magnitude of a unit vector equals one: ‖u‖ = 1.
  • Normalization preserves direction: u is collinear with the original vector v.
  • Unit vectors form the unit circle (2D) or unit sphere (3D) under the Euclidean norm.

Notation varies: standard basis unit vectors in Cartesian coordinates are often written i, j, k or êx, êy, êz. More generally, a unit vector may be written with a hat, as in v̂, or using boldface with an overbar in some texts. See also vector for background on vectors and norms.

History and context. The idea of normalizing a vector arose naturally with the development of vector algebra and analytic geometry in the 19th and early 20th centuries as mathematicians and physicists formalized directional quantities. Unit vectors provide a compact way to describe orientation independent of scale, which is useful in coordinate systems and when defining directional derivatives, directional cosines, and basis elements.

Applications and computational notes. Unit vectors appear across disciplines: they define axes in coordinate systems, specify directions of force or velocity in physics, serve as normals for surfaces in geometry and graphics, and are essential in algorithms that require direction only (e.g., ray tracing, gradient descent directions). Computationally, care is required when normalizing very small or very large vectors because of underflow or overflow; common approaches include scaling the vector before dividing or using robust library routines. Algorithms such as Gram–Schmidt produce orthonormal sets—collections of mutually perpendicular unit vectors—used to build coordinate bases.

In summary, a unit vector is the normalized form of a nonzero vector and is a basic tool for representing pure direction. The process of dividing by a norm gives a simple, reversible relationship between a vector and its unit form when the original vector is nonzero.