In mathematics, more specifically general topology, the divisor topology is a specific topology on the set
of positive integers greater than or equal to two. The divisor topology is the poset topology for the partial order relation of divisibility of integers on
.
Construction
The sets
for
form a basis for the divisor topology[1] on
, where the notation
means
is a divisor of
.
The open sets in this topology are the lower sets for the partial order defined by
if
. The closed sets are the upper sets for this partial order.
Properties
All the properties below are proved in [1] or follow directly from the definitions.
- The closure of a point
is the set of all multiples of
.
- Given a point
, there is a smallest neighborhood of
, namely the basic open set
of divisors of
. So the divisor topology is an Alexandrov topology.
is a T0 space. Indeed, given two points
and
with
, the open neighborhood
of
does not contain
.
is a not a T1 space, as no point is closed. Consequently,
is not Hausdorff.
- The isolated points of
are the prime numbers.
- The set of prime numbers is dense in
. In fact, every dense open set must include every prime, and therefore
is a Baire space.
is second-countable.
is ultraconnected, since the closures of the singletons
and
contain the product
as a common element.
- Hence
is a normal space. But
is not completely normal. For example, the singletons
and
are separated sets (6 is not a multiple of 4 and 4 is not a multiple of 6), but have no disjoint open neighborhoods, as their smallest respective open neighborhoods meet non-trivially in
.
is not a regular space, as a basic neighborhood
is finite, but the closure of a point is infinite.
is connected, locally connected, path connected and locally path connected.
is a scattered space, as each nonempty subset has a first element, which is an isolated element of the set.
- The compact subsets of
are the finite subsets, since any set
is covered by the collection of all basic open sets
, which are each finite, and if
is covered by only finitely many of them, it must itself be finite. In particular,
is not compact.
is locally compact in the sense that each point has a compact neighborhood (
is finite). But points don't have closed compact neighborhoods (
is not locally relatively compact.)
References
- ^ a b Steen & Seebach, example 57, p. 79-80