A well-defined function must map every element of the domain to an element of its codomain. So, for example, the function:
has no valid value for f(0). It is thus not a function on the set R of real numbers; R can't be its domain. It is usually either defined as a function on R \ {0}, or the "gap" is plugged by specifically defining f(0); for example:
The domain of given function can be restricted to a subset. Suppose that g: A → B, and S ⊆ A. Then the restriction of g to S is written:
See also: Function codomain
wikipedia.org dumped 2003-03-17 with terodump