Quire

Comparability

From Quire, the open reference·Property of elements related by inequalities·2 min read
Hasse diagram of the natural numbers, partially ordered by "x≤y if x divides y". The numbers 4 and 6 are incomparable, since neither divides the other.
Hasse diagram of the natural numbers, partially ordered by "x≤y if x divides y". The numbers 4 and 6 are incomparable, since neither divides the other. Image: source and license

In mathematics, two elements x and y of a set P are said to be comparable with respect to a binary relation ≤ if at least one of x ≤ y or y ≤ x is true. They are called incomparable if they are not comparable.

Rigorous definition#edit

A binary relation on a set {\displaystyle P} is by definition any subset {\displaystyle R} of {\displaystyle P\times P.} Given {\displaystyle x,y\in P,} {\displaystyle xRy} is written if and only if {\displaystyle (x,y)\in R,} in which case {\displaystyle x} is said to be related to {\displaystyle y} by {\displaystyle R.} An element {\displaystyle x\in P} is said to be {\displaystyle R}-comparable, or comparable (with respect to {\displaystyle R}), to an element {\displaystyle y\in P} if {\displaystyle xRy} or {\displaystyle yRx.} Often, a symbol indicating comparison, such as {\displaystyle <} (or {\displaystyle \leq ,} {\displaystyle >,} {\displaystyle \geq ,} and many others) is used instead of {\displaystyle R,} in which case {\displaystyle x<y} is written in place of {\displaystyle xRy,} which is why the term "comparable" is used.

Comparability with respect to {\displaystyle R} induces a canonical binary relation on {\displaystyle P}; specifically, the comparability relation induced by {\displaystyle R} is defined to be the set of all pairs {\displaystyle (x,y)\in P\times P} such that {\displaystyle x} is comparable to {\displaystyle y}; that is, such that at least one of {\displaystyle xRy} and {\displaystyle yRx} is true. Similarly, the incomparability relation on {\displaystyle P} induced by {\displaystyle R} is defined to be the set of all pairs {\displaystyle (x,y)\in P\times P} such that {\displaystyle x} is incomparable to {\displaystyle y;} that is, such that neither {\displaystyle xRy} nor {\displaystyle yRx} is true.

If the symbol {\displaystyle <} is used in place of {\displaystyle \leq } then comparability with respect to {\displaystyle <} is sometimes denoted by the symbol {\displaystyle {\overset {<}{\underset {>}{=, and incomparability by the symbol {\displaystyle {\cancel {\overset {<}{\underset {>}{=\!}.[1] Thus, for any two elements {\displaystyle x} and {\displaystyle y} of a partially ordered set, exactly one of {\displaystyle x\ {\overset {<}{\underset {>}{=\ y} and {\displaystyle x{\cancel {\overset {<}{\underset {>}{=y} is true.

Example#edit

A totally ordered set is a partially ordered set in which any two elements are comparable. The Szpilrajn extension theorem states that every partial order is contained in a total order. Intuitively, the theorem says that any method of comparing elements that leaves some pairs incomparable can be extended in such a way that every pair becomes comparable.

Properties#edit

Both of the relations comparability and incomparability are symmetric, that is {\displaystyle x} is comparable to {\displaystyle y} if and only if {\displaystyle y} is comparable to {\displaystyle x,} and likewise for incomparability.

Comparability graphs#edit

The comparability graph of a partially ordered set {\displaystyle P} has as vertices the elements of {\displaystyle P} and has as edges precisely those pairs {\displaystyle \{x,y\ of elements for which {\displaystyle x\ {\overset {<}{\underset {>}{=\ y}.[2]

Classification#edit

When classifying mathematical objects (e.g., topological spaces), two criteria are said to be comparable when the objects that obey one criterion constitute a subset of the objects that obey the other, which is to say when they are comparable under the partial order ⊂. For example, the T1 and T2 criteria are comparable, while the T1 and sobriety criteria are not.

See also#edit

References#edit

  1. ^ Trotter, William T. (1992), Combinatorics and Partially Ordered Sets:Dimension Theory, Johns Hopkins Univ. Press, p. 3
  2. ^ Gilmore, P. C.; Hoffman, A. J. (1964), "A characterization of comparability graphs and of interval graphs", Canadian Journal of Mathematics, 16: 539–548, doi:10.4153/CJM-1964-055-5.

This article is mirrored from “Comparability” on English Wikipedia, revision 1360404613, written by Wikipedia's contributors and licensed CC BY-SA 4.0. Editing it here creates a Quire copy with its own history.

Synced with Wikipedia revision 1360404613 · just now