Degeneracy (mathematics)

This article is about degeneracy in mathematics. For the degeneracy of a graph, see degeneracy (graph theory). For other uses, see Degeneracy (disambiguation).

In mathematics, a degenerate case is a limiting case in which an element of a class of objects is qualitatively different from the rest of the class and hence belongs to another, usually simpler, class. Degeneracy is the condition of being a degenerate case.

The definitions of many classes of composite or structured objects include (often implicitly) inequalities. For example, the angles and the side lengths of a triangle are supposed to be positive. The limiting cases, where one of these elements of the triangle is zero, are degenerate triangles.

Often, the degenerate cases are the exceptional cases where changes to the usual dimension or the cardinality of the object (or of some part of it) occur. For example, a triangle is an object of dimension two, and a degenerate triangle is contained in a line, and its dimension is thus one. Similarly, the solution set of a system of equations that depends on parameters generally has a fixed cardinality and dimension, but cardinality and/or dimension may be different for some exceptional values, called degenerate cases. In such a degenerate case, the solution set is said to be degenerate.

For some classes of composite objects, the degenerate cases depend on the properties that are specifically studied. In particular, the class of objects may often be defined or characterized by systems of equations. Commonly, a given class of objects may be defined by several different systems of equations, and these different systems of equations may lead to different degenerate cases, while characterizing the same non-degenerate cases. This may be the reason for which there is no general definition of degeneracy, although the concept is widely used, and defined, if needed, in each specific situation.

A degenerate case thus has special features, which makes it non-generic. However not all non-generic cases are degenerate. For example, right triangles, isosceles triangles and equilateral triangles are non-generic and non-degenerate. Frequently, degenerate cases correspond to singularities either in the object or in some configuration space. For example, a conic section is degenerate if and only if it has singular points.

In geometry

Conic section

Main article: Degenerate conic

A degenerate conic is a conic section (a second-degree plane curve, defined by a polynomial equation of degree two) that fails to be an irreducible curve.

• A point is a degenerate circle, namely one with radius 0.
• The line is a degenerate case of a parabola if the parabola resides on a tangent plane. In inversive geometry, a line is a degenerate case of a circle, with infinite radius.
• A line segment can be viewed as a degenerate case of an ellipse in which the semiminor axis goes to zero, the foci go to the endpoints, and the eccentricity goes to one.
• An ellipse can also degenerate into a single point.
• A hyperbola can degenerate into two lines crossing at a point, through a family of hyperbolae having those lines as common asymptotes.

Triangle

• A degenerate triangle has collinear vertices and zero area, and thus coincides with a segment covered twice.

Rectangle

• A segment is a degenerate case of a rectangle, if this has a side of length 0.

• For any non-empty subset $S \subseteq \{1, 2, \ldots, n\}$, there is a bounded, axis-aligned degenerate rectangle

$R \triangleq \left\{\mathbf{x} \in \mathbb{R}^n: x_i = c_i \ (\text{for } i\in S) \text{ and } a_i \leq x_i \leq b_i \ (\text{for } i \notin S)\right\}$

where $\mathbf{x} \triangleq [x_1, x_2, \ldots, x_n]$ and $a_i, b_i, c_i$ are constant (with $a_i \leq b_i$ for all $i$). The number of degenerate sides of $R$ is the number of elements of the subset $S$. Thus, there may be as few as one degenerate "side" or as many as $n$ (in which case $R$ reduces to a singleton point).

Convex polygon

• A convex polygon is degenerate if (at least) two consecutive sides are aligned or some sides have a zero length. Thus a degenerate convex polygon of n sides looks like a polygon with fewer sides. In the case of triangles, this definition coincide with the one that has been given above.

Convex polyhedron

• A convex polyhedron is degenerate if either two adjacent facets are coplanar or two edges are aligned. In the case of a tetrahedron, this is equivalent to say that all of its vertices lie in the same plane, giving it zero volume.

Standard torus

• A sphere is a degenerate standard torus where the axis of revolution passes through the center of the generating circle, rather than outside it.

Sphere

• When the radius of a sphere goes to zero, the resulting degenerate sphere of zero volume is a point.

Other

• See general position for other examples.

Elsewhere

• A set containing a single point is a degenerate continuum.
• Objects such as the digon and monogon can be viewed as degenerate cases of polygons: valid in a general abstract mathematical sense, but not part of the original Euclidean conception of polygons.

• A random variable which can only take one value has a degenerate distribution; if that value is the real number 0, its probability density is the Dirac delta function.

• Similarly, roots of a polynomial are said to be degenerate if they coincide, since generically the n roots of an nth degree polynomial are all distinct. This usage carries over to eigenproblems: a degenerate eigenvalue (i.e. a multiple coinciding root of the characteristic polynomial) is one that has more than one linearly independent eigenvector.

• In quantum mechanics any such multiplicity in the eigenvalues of the Hamiltonian operator gives rise to degenerate energy levels. Usually any such degeneracy indicates some underlying symmetry in the system.

See also

• Degeneracy (graph theory)
• Degenerate form
• Trivial (mathematics)
• Pathological (mathematics)
• Vacuous truth

External links

Weisstein, Eric W., "Degenerate", MathWorld.