Overview

Every vector space has a canonical affine algebraic variety structure. This makes sense intuitively since (informally) each vector space is somehow linear… and thus affine.

Affine algebraic variety structure

Let be a finite-dimensional vector space. Given an ordered basis of , consider the vector space isomorphism

The set has a unique affine algebraic variety structure that makes an isomorphism of affine algebraic varieties for all ordered bases of .

Proof

Throughout this proof, to make things easier to write, assume for some .

In order to prove the statement above, we can show that we can put an affine algebraic variety structure on V, then we must show that it is unique.

Existence of affine variety structure

To give V an affine algebraic variety structure, first we need to make sure is a homeomorphism. Therefore, for every closed set , let be a closed subset of .

Now, we need to define . Since we have a bijection between and , we can use that to build .

For let be defined as the map .

Therefore, we have the pair such that and

where

This means has an affine algebraic variety structure.

Uniqueness

Using this lemma, given two ordered basis and for V, it suffices to show that is an isomorphism where denotes the affine algebraic variety structure inherited from .

We can illustrate this with the commutative diagram below:

We know that the diagram commutes, therefore

We know by construction that and are isomorphisms. However, we need to show that the middle term is an isomorphism. It is apparent that it is a vector space isomorphism, but not that it is a morphism of affine algebraic varieties.


Lemma For the map

is a morphism of affine algebraic varieties.

Proof Fix .

Therefore

Now for an arbitrary ,

Since we know that is an algebra homomorphism,

which is a product of polynomials, and is thus a polynomials. Stated another way, .

Now the only thing that remains is to prove that is continuous.

Let be closed. Thus

Consider the preimage of the set :

Thus is closed in and is continuous.

With both of these conditions, the map which is left multiplication by a matrix is a morphism.


By this lemma, is a morphism because it can be written as the matrix representing the change of basis. Thus, is a composition of morphisms so it must be a morphism. Note this same process can be done with , so is a morphism with an inverse morphism so it is an isomorphism. By the first lemma, for any choice of ordered basis on the two affine algebraic variety structures are the same.

Linear maps are morphisms

Suppose that is another finite-dimensional vector space. Equip the sets and with the affine algebraic variety structures resulting from Part (i), and let be a linear map. Prove that is a morphism of affine algebraic varieties.

For simplicity let .

Finite dimensional vector spaces are isomorphic to . Thus, we know we can find a matrix representation of such that the following diagram commutes for a given choice of ordered basis for W.

Therefore, we see that the function can be written as

Similarly to above, we know that and are isomorphism by construction. By the lemma above, we know that the map for left multplication by a matrix (in this case ) is a morphism, thus is a composition of morphism and thus is a morphism.