If is a product of rings, then for every
i in
I we have a
surjective ring homomorphism which projects the product on the
ith coordinate. The product
R together with the projections
pi has the following
universal property: :if
S is any ring and is a ring homomorphism for every
i in
I, then there exists
precisely one ring homomorphism such that for every
i in
I. This shows that the product of rings is an instance of
products in the sense of category theory. When
I is finite, the underlying additive group of coincides with the
direct sum of the additive groups of the
Ri. In this case, some authors call
R the "direct sum of the rings
Ri" and write , but this is incorrect from the point of view of
category theory, since it is usually not a
coproduct in the
category of rings (with identity): for example, when two or more of the
Ri are non-
trivial, the inclusion map fails to map 1 to 1 and hence is not a ring homomorphism. (A finite coproduct in the
category of
commutative algebras over a commutative ring is a
tensor product of algebras. A coproduct in the category of algebras is a
free product of algebras.) Direct products are commutative and associative
up to natural isomorphism, meaning that it doesn't matter in which order one forms the direct product. If
Ai is an
ideal of
Ri for each
i in
I, then is an ideal of
R. If
I is finite, then the
converse is true, i.e., every ideal of
R is of this form. However, if
I is infinite and the rings
Ri are non-trivial, then the converse is false: the set of elements with all but finitely many nonzero coordinates forms an ideal which is not a direct product of ideals of the
Ri. The ideal
A is a
prime ideal in
R if all but one of the
Ai are equal to
Ri and the remaining
Ai is a prime ideal in
Ri. However, the converse is not true when
I is infinite. For example, the
direct sum of the
Ri form an ideal not contained in any such
A, but the
axiom of choice gives that it is contained in some
maximal ideal which is
a fortiori prime. An element
x in
R is a
unit if and only if all of its components are units, i.e., if and only if
pi(
x) is a unit in
Ri for every
i in
I. The
group of units of
R is the
product of the groups of units of the
Ri. A product of two or more non-trivial rings always has nonzero
zero divisors: if
x is an element of the product whose coordinates are all zero except
pi(
x) and
y is an element of the product with all coordinates zero except
pj(
y) where
i ≠
j, then
xy = 0 in the product ring. ==References==