linear extension
Let be a commutative ring, a free -module, a basis of, and a further -module. Each element then has aunique representation
where for all , and only finitely many arenon-zero. Given a set map we may therefore define the -module homomorphism , called the linear extension of , such that
The map is the unique homomorphism from to whose restriction
to is .
The above observation has a convenient reformulation in terms of category theory. Let denote the category
of -modules, and the category of sets. Consider the adjoint functors
, the forgetful functor
that maps an -module to its underlying set, and ,the free module
functor
that maps a set to the free -module generated by that set. To say that is right-adjoint to is the same as saying that every set map from to , the set underlying , corresponds naturally and bijectively to an -module homomorphism from to .
Similarly, given a map , we may define thebilinear extension
which is the unique bilinear map from to whose restrictionto is .
Generally, for any positive integer and a map , we may define the -linear extension
quite compactly using multi-index notation: .
Usage
The notion of linear extension is typically used as amanner-of-speaking. Thus, when a multilinear map is definedexplicitly in a mathematical text, the images of the basis elementsare given accompanied by the phrase “by multilinear extension” orsimilar.