Dual number

From Wikipedia, the free encyclopedia
Jump to: navigation, search

In linear algebra, the dual numbers extend the real numbers by adjoining one new element ε with the property ε2 = 0 (ε is nilpotent). The collection of dual numbers forms a particular two-dimensional commutative unital associative algebra over the real numbers. Every dual number has the form z = a + bε with a and b uniquely determined real numbers.

The algebra of dual numbers is a ring that is a local ring since the principal ideal generated by ε is its only maximal ideal. Dual numbers form the coefficients of dual quaternions.

Linear representation[edit]

Using matrices, dual numbers can be represented as

\varepsilon=\begin{pmatrix}0 & 1 \\0 & 0 \end{pmatrix}\quad\text{and}\quad a + b\varepsilon = \begin{pmatrix}a & b \\ 0 & a \end{pmatrix}.

The sum and product of dual numbers are then calculated with ordinary matrix addition and matrix multiplication; both operations are commutative and associative within the algebra of dual numbers.

This correspondence is analogous to the usual matrix representation of complex numbers. However, it is not the only representation with 2 × 2 real matrices, as is shown in the profile of 2 × 2 real matrices. Like the complex plane and split-complex number plane, the dual numbers are one of the realizations of planar algebra.


The "unit circle" of dual numbers consists of those with a = 1 or −1 since these satisfy z z* = 1 where z* = abε. However, note that

 \exp(b \varepsilon) = \left(\sum^\infty_{n=0} (b\varepsilon)^n / n!\right) = 1 + b \varepsilon \!,

so the exponential mapping applied to the ε-axis covers only half the "circle".

If a ≠ 0 and m = b /a, then z = a(1 + m ε) is the polar decomposition of the dual number z, and the slope m is its angular part. The concept of a rotation in the dual number plane is equivalent to a vertical shear mapping since (1 + p ε)(1 + q ε) = 1 + (p+q) ε.

The dual number plane is used to represent the naive spacetime of Galileo in a study called Galilean invariance since the classical event-transformation with velocity v is expressed by a shear mapping:

(t',x') = (t,x)\begin{pmatrix}1 & v \\0 & 1 \end{pmatrix}, that is \ \ t'=t,\ \  x' = vt + x \!.


Given two dual numbers p, and q, they determine the set of z such that the difference in slopes ("Galilean angle") between the lines from z to p and q is constant. This set is a cycle in the dual number plane; since the equation setting the difference in slopes of the lines to a constant is a quadratic equation in the real part of z, a cycle is a parabola. The "cyclic rotation" of the dual number plane occurs as a motion of the projective line over dual numbers. According to Yaglom (pp. 92,3), the cycle Z = {z : y = α x2} is invariant under the composition of the shear

x_1 = x ,\ \  y_1 = vx + y \ \ with the translation
x' = x_1 = v/2a  ,\ \  y' = y_1 + v^2/4a \ .

This composition is a cyclic rotation; the concept has been further developed by V. V. Kisil.[1]

Algebraic properties[edit]

In abstract algebra terms, the dual numbers can be described as the quotient of the polynomial ring R[X] by the ideal generated by the polynomial X2,


The image of X in the quotient is the unit ε. With this description, it is clear that the dual numbers form a commutative ring with characteristic 0. The inherited multiplication gives the dual numbers the structure of a commutative and associative algebra over the reals of dimension two. The algebra is not a division algebra or field since the elements of the form 0 + bε are not invertible. All elements of this form are zero divisors (also see the section "Division"). The algebra of dual numbers is isomorphic to the exterior algebra of \mathbb{R}^1.


This construction can be carried out more generally: for a commutative ring R one can define the dual numbers over R as the quotient of the polynomial ring R[X] by the ideal (X2): the image of X then has square equal to zero and corresponds to the element ε from above.

This ring and its generalisations play an important part in the algebraic theory of derivations and Kähler differentials (purely algebraic differential forms).

Over any ring R, the dual number a + bε is a unit (i.e. multiplicatively invertible) if and only if a is a unit in R. In this case, the inverse of a + bε is a−1ba−2ε. As a consequence, we see that the dual numbers over any field (or any commutative local ring) form a local ring, its maximal ideal being the principal ideal generated by ε.


One application of dual numbers is automatic differentiation. Consider the real dual numbers above. Given any real polynomial P(x) = p0+p1x+p2x2+...+pnxn, it is straightforward to extend the domain of this polynomial from the reals to the dual numbers. Then we have this result:

P(a+b\varepsilon)=&p_0 + p_1(a + b\varepsilon) + \ldots + p_n(a + b\varepsilon)^n\\
=&p_0 + p_1 a + p_2 a^2 + \ldots + p_n a^n\\
& + p_1 b\varepsilon + 2 p_2 a b\varepsilon + \ldots + n p_n a^{n-1} b\varepsilon\\

where P^\prime is the derivative of P.[2]

By computing over the dual numbers, rather than over the reals, we can use this to compute derivatives of polynomials.

More generally, we can extend any (smooth) real function to the dual numbers by looking at its Taylor series: f(a+b\varepsilon)=\sum_{n=0}^{\infty} {{f^{(n)} (a)b^n \varepsilon^n} \over {n!}}=f(a)+bf'(a)\varepsilon.
By computing compositions of these functions over the dual numbers and examining the coefficient of ε in the result we find we have automatically computed the derivative of the composition.

This effect can be explained from the non-standard analysis viewpoint. The unit ε of dual numbers is a close relative to infinitesimal used in non-standard calculus: indeed the square (or any higher power) of ε is exactly zero and the square of an infinitesimal is almost zero at this infinitesimal's scale (is an infinitesimal of a higher order more precisely).


Dual numbers find applications in physics, where they constitute one of the simplest non-trivial examples of a superspace. The direction along ε is termed the "fermionic" direction, and the real component is termed the "bosonic" direction. The fermionic direction earns this name from the fact that fermions obey the Pauli exclusion principle: under the exchange of coordinates, the quantum mechanical wave function changes sign, and thus vanishes if two coordinates are brought together; this physical idea is captured by the algebraic relation ε2 = 0.


Division of dual numbers is defined when the real part of the denominator is non-zero. The division process is analogous to complex division in that the denominator is multiplied by its conjugate in order to cancel the non-real parts.

Therefore, to divide an equation of the form:

{a+b\varepsilon \over c+d\varepsilon}

We multiply the top and bottom by the conjugate of the denominator:

= {(a+b\varepsilon)(c-d\varepsilon) \over (c+d\varepsilon)(c-d\varepsilon)}
= {ac-ad\varepsilon+bc\varepsilon-bd\varepsilon^2 \over (c^2+cd\varepsilon-cd\varepsilon-d^2\varepsilon^2)}
= {ac-ad\varepsilon+bc\varepsilon-0 \over c^2-0}
= {ac + \varepsilon(bc - ad) \over c^2}
= {a \over c} + {(bc - ad) \over c^2}\varepsilon

Which is defined when c is non-zero.

If, on the other hand, c is zero while d is not, then the equation

{a+b\varepsilon = (x+y\varepsilon) d\varepsilon} = {xd\varepsilon + 0}
  1. has no solution if a is nonzero
  2. is otherwise solved by any dual number of the form
{b \over d} + {y\varepsilon}.

This means that the non-real part of the "quotient" is arbitrary and division is therefore not defined for purely nonreal dual numbers. Indeed, they are (trivially) zero divisors and clearly form an ideal of the associative algebra (and thus ring) of the dual numbers.

See also[edit]

Notes and references[edit]

  1. ^ V.V. Kisil (2007) "Inventing a Wheel, the Parabolic One" arXiv:0707.4024
  2. ^ Berland, Håvard. "Automatic differentiation". Retrieved 13 May 2013.