Isolated point

A set that is made up only of isolated points is called a discrete set (see also discrete space). Any discrete subset S of Euclidean space must be countable, since the isolation of each of its points together with the fact the rationals are dense in the reals means that the points of S may be mapped into a set of points with rational coordinates, of which there are only countably many. However, not every countable set is discrete, of which the rational numbers under the usual Euclidean metric are the canonical example.

A set with no isolated point is said to be dense-in-itself (every neighbourhood of a point contains other points of the set). A closed set with no isolated point is called a perfect set (it has all its limit points and none of them are isolated from it).

The number of isolated points is a topological invariant, i.e. if two topological spaces ${\displaystyle X}$ and ${\displaystyle Y}$ are homeomorphic, the number of isolated points in each is equal.

Topological spaces in the following examples are considered as subspaces of the real line with the standard topology.

• For the set ${\displaystyle S=\{0\}\cup [1,2]}$, the point 0 is an isolated point.
• For the set ${\displaystyle S=\{0\}\cup \{1,1/2,1/3,\dots \}}$, each of the points 1/k is an isolated point, but 0 is not an isolated point because there are other points in S as close to 0 as desired.
• The set ${\displaystyle {\mathbb {N} }=\{0,1,2,\ldots \}}$ of natural numbers is a discrete set.
• The Morse lemma states that non-degenerate critical points of certain functions are isolated.

Consider the set ${\displaystyle F}$ of points ${\displaystyle x}$ in the real interval ${\displaystyle (0,1)}$ such that every digit of their binary representation ${\displaystyle x_{i}}$ fulfills the following conditions:

• Either ${\displaystyle x_{i}=0}$ or ${\displaystyle x_{i}=1}$.
• ${\displaystyle x_{i}=1}$ only for finitely many indices ${\displaystyle i}$.
• If ${\displaystyle m}$ denotes the biggest index such that ${\displaystyle x_{m}=1}$, then ${\displaystyle x_{m-1}=0}$.
• If ${\displaystyle x_{i}=1}$ and ${\displaystyle i<m}$, then exactly one of the following two conditions holds: ${\displaystyle x_{i-1}=1}$, ${\displaystyle x_{i+1}=1}$. Informally, this condition means that every digit of the binary representation of ${\displaystyle x}$ which equals to 1 belongs to a pair ...0110..., except for ...010... at the very end.

Now, ${\displaystyle F}$ is an explicit set consisting entirely of isolated points[1] which has the counter-intuitive property that its closure is an uncountable set.[2]

Another set ${\displaystyle F}$ with the same properties can be obtained as follows. Let ${\displaystyle C}$ be the middle-thirds Cantor set, let ${\displaystyle I_{1},I_{2},I_{3},\ldots }$ be the component intervals of ${\displaystyle [0,1]-C}$, and let ${\displaystyle F}$ be a set consisting of one point from each ${\displaystyle I_{k}}$. Since each ${\displaystyle I_{k}}$ contains only one point from ${\displaystyle F}$, every point of ${\displaystyle F}$ is an isolated point. However, if ${\displaystyle p}$ is any point in the Cantor set, then every neighborhood of ${\displaystyle p}$ contains at least one ${\displaystyle I_{k}}$, and hence at least one point of ${\displaystyle F}$. It follows that each point of the Cantor set lies in the closure of ${\displaystyle F}$, and therefore ${\displaystyle F}$ has uncountable closure.

• Acnode
• Adherent point
• Limit point

• Gomez-Ramirez, Danny (2007), "An explicit set of isolated points in R with uncountable closure", Matemáticas: Enseñanza universitaria, Escuela Regional de Matemáticas. Universidad del Valle, Colombia, 15: 145–147

• Weisstein, Eric W. "Isolated Point". MathWorld.