17.2: Operations on Relations

- Last updated
-
Feb 19, 2022
-
Save as PDF
-
Viewing relations as subsets of Cartesian products suggests ways to build new relations from old.
Definition: Union (of relations )
the relation where means that at least one of or is true
Definition: Intersection (of relations )
the relation where means that both and are true
Definition: Complement (of relation )
the relation where means that is not true
Definition:
alternative notation for
Note
Considering relations as subsets of Cartesian products, the above relation operations mean precisely the same thing as the corresponding set operations.
Example : Union of “less than” and “equal to” relations.
Consider the relations and on and let be the union Then means that at least one of or is true. That is, is the same as the relation
Example : Sibling relations.
Let represent the set of all living humans. Let relations be defined by
- if have the same father; and
- if have the same mother.
Set Then means that have the same parents.
Example : Complement of the subset relation.
Let be a universal set and consider the relation on Then means that is not a subset of which can only happen if some elements of are not in In other words, means that
- Careful.
-
Relation does not (necessarily) mean Draw a representative Venn diagram to see why.
Unlike functions, which can only be reversed if bijective, every relation can be reversed by simply stating the relationship in the reverse order.
Definition: Inverse (of a relation )
the relation where means that is true
Note
- As subsets of Cartesian products, if then and if and only if
- A relation and its inverse express the same relationship between elements of two sets and just phrased in the opposite order. In logical terms,
Example : Parent/child relations.
Let represent the set of all living humans, and let represent the relation on where means human is the parent of human Then means human is the child of human Both relations express the same information, but in a different order.
Example : Inverse of Division Relation.
Recall that is a relation on where means that divides Then for the inverse relation, means is a multiple of Both relations express the same information, but in a different order.
Example : Inverse of Logical Equivalence
Let represent the set of all possible logical statements. We have a relation on where means that logical statement involves the same statement variables and has the same truth table as logical statement Since if and only if we conclude that the logical equivalence relation on is its own inverse.
There are two more set-theoretic ideas we can reinterpret as relations.
Definition: Empty Relation
the relation between sets and corresponding to the empty subset so that is always false
Definition: Universal Relation
the relation between sets and corresponding to the full subset so that is always true.