Chapter 3

Basic Topology of R

3.1. Open and Closed Sets

We have already defined the -neighborhood V (a) of a number a ∈ R; namely, V (a) =
(a − , a + ), an open interval around a.

Definition 3.1. (i) Let a ∈ A. We say a is an interior point of A if ∃ > 0 (V (a) ⊆ A).
(ii) A set O ⊆ R is said to be open if every element a ∈ O is an interior point of O.

Remark 3.2. Open intervals (a, b), (a, ∞) and (−∞, b) are open sets in R. However,
intervals of the form [a, b], (a, b] or [a, b) are not open sets. Try to prove these statements
yourselves.
S
Theorem 3.1. (i) If Oα ⊆ R is open for each α ∈ I, so is the union α∈I Oα .
(ii) If O1 , O2 , · · · , On are open sets, so is the intersection ni=1 Oi .
T

S
Proof. To prove (i), let O = α∈I Oα . Take any point a ∈ O. Then a ∈ Oα for some α ∈ I.
For this α, since Oα is open, there exists a neighborhood V (a) ⊆ Oα since a ∈ Oα . Clearly
this neighborhood V (a) is also contained in the union O. This proves O is open.
For (ii), let O = ni=1 Oi . Let a ∈ O. Then a ∈ Oi for each i = 1, 2, · · · , n. Since
T
Oi is open, there exists a neighborhood Vi (a) ⊆ Oi for i = 1, 2, · · · , n, where i > 0.
Let  = min{1 , 2 , · · · , n }. Then  > 0 and V (a) ⊆ Vi (a). Hence V (a) ⊆ Oi for all
i = 1, 2, · · · , n. So V (a) ⊆ ∩ni=1 Oi = O. Hence O is open. 

Remark 3.3. Part (ii) of Theorem 3.1 may be false for infinite number of open sets. For
example,
∞  
\ 1 1
−1 − , 1 + = [−1, 1].
i i
i=1

Definition 3.4. (i) A point x is called a limit point of a set A if ∀  > 0 (V̂ (x) ∩ A 6= ∅),
where V̂ (x) = V (x) \ {x} is called the punctured -neighborhood of x.
The set of all limit points of A will be denoted by L(A).
(ii) A point a in A is called an isolated point of A if a ∈
/ L(A).
(iii) A set F is called a closed set if L(F ) ⊆ F.

1
2 3. Basic Topology of R

Theorem 3.2. A point x is a limit point of set A if and only if there exists a sequence (an )
in A such that an 6= x for all n ∈ N and (an ) → x.

Proof. Assume x ∈ L(A). Then ∀ n ∈ N ∃ an ∈ V̂1/n (x) ∩ A. Hence 0 < |an − x| < 1/n; so
the sequence (an ) has the required property.
Now assume (an ) is a sequence such that an ∈ A, an 6= x for all n ∈ N and (an ) → x.
Given any  > 0, there exists an N ∈ N such that |aN − x| < ; hence aN ∈ V (x). Clearly
aN 6= x; so aN ∈ V̂ (x) ∩ A and thus, by the definition, x ∈ L(A). 
Theorem 3.3. A set F is closed if and only if the limit of every Cauchy sequence (or
convergent sequence) in F belongs to F .

Proof. 1. Let F be closed. Let (xn ) be a Cauchy sequence with xn ∈ F. By the CC,
(xn ) → x. We show x ∈ F. Suppose not: x ∈ / F. Then xn 6= x for all n ∈ N. By the theorem
above, x is a limit point of F and hence x ∈ F, a contradiction. So x ∈ F.
2. Now assume that the limit of every Cauchy sequence in F belongs to F . We show
that F is closed. Let x ∈ L(F ). Then, by the theorem above, there exists a sequence (xn )
with xn ∈ F , xn 6= x, such that (xn ) → x. This implies (xn ) is a Cauchy sequence in F ;
hence x ∈ F. So L(F ) ⊆ F and F is closed. 
Example 3.1. (i) Each element in the set A = { n1 : n ∈ N} is an isolated point of A. Also
0 is the only limit point of A. Since 0 ∈
/ A, this set A is not closed.
(ii) Closed intervals [a, ∞), (−∞, b] and [a, b] are closed sets. (Exercise.)
(iii) The interval [a, b) = {x ∈ R : a ≤ x < b} is neither open nor closed.
(iv) Every x ∈ R is a limit point of Q; this follows from the density of Q in R.
Definition 3.5. The closure of a set A, denoted by Ā, is defined to be the union of A and
L(A); namely Ā = A ∪ L(A).
Theorem 3.4. For any set A, the closure Ā is a closed set and is the smallest closed set
containing A.

Proof. 1. We first prove Ā is closed. Let a be a limit point of Ā. We show a ∈ Ā. If a ∈ A
then a ∈ Ā. So assume a ∈ / A. Since a is a limit point of Ā, there exists a sequence (xn )
with (xn ) → a and xn ∈ Ā and xn 6= a for all n ∈ N. For any n ∈ N, if xn ∈ A define
yn = xn and hence yn 6= a; if xn ∈ / A, since xn ∈ Ā, then xn ∈ L(A), and in this case,
define yn ∈ A such that 0 < |yn − xn | < |xn − a| and hence yn 6= a; such a yn exists from
the definition of limit point xn with  = |xn − a| > 0. Therefore, we obtain a sequence (yn )
with the property: yn ∈ A, yn 6= a and
|yn − a| ≤ |yn − xn | + |xn − a| ≤ 2|xn − a| ∀ n ∈ N.
Hence yn ∈ A, (yn ) → a and yn 6= a for all n ∈ N. By the theorem above, this shows that
a is a limit point of A; hence a ∈ Ā. We have proved that Ā is closed.
2. Clearly Ā contains A. To show that Ā is the smallest closed set containing A, assume
B is any closed set containing A and we want to show Ā ⊆ B. Let x ∈ Ā and we show
x ∈ B. If x ∈ A then x ∈ B. Assume x ∈ L(A). Then ∃ xn ∈ A, xn 6= x such that (xn ) → x.
Since xn ∈ B, this shows that x is also a limit point of B (this actually shows that if A ⊆ B,
then L(A) ⊆ L(B)). Since B is closed, we have x ∈ B. So Ā ⊆ B. 
Corollary 3.5. If A ⊆ B, then Ā ⊆ B̄.

Proof. Let A ⊆ B. Then A ⊆ B̄. Since B̄ is closed, by Theorem 3.4, Ā ⊆ B̄. 

3.2. Compact Sets 3

Complements. Given a set A ⊆ R, the complement Ac of A in R is defined by

Ac = R \ A = {x ∈ R : x ∈
/ A}.

The result of the following theorem can be used as a definition for closed sets.
Theorem 3.6. A set in R is open if and only if its complement is closed. Likewise, a set
in R is closed if and only if its complement is open.

Proof. 1. Assume O is open; we show that F = Oc is closed. Let x ∈ L(F ). Suppose

x 6∈ F. Then x ∈ O and hence V (x) ⊂ O for some  > 0. This implies that V (x) ∩ F = ∅,
contradicting x ∈ L(F ).
2. Assume F is closed; we show that O = F c is open. Let a ∈ O; then a 6∈ F. Since
F is closed, a 6∈ L(F ); hence ∃  > 0 such that V̂ (a) ∩ F = ∅. Since a 6∈ F , it follows that
V (a) ∩ F = ∅ and thus V (a) ⊂ F c = O; hence, by definition, O is open.
3. Since (Ac )c = A for all sets in R, it is easily seen that the two steps above complete
the proof. 
Theorem 3.7. (i) The union of a finite collection of closed sets is closed.
(ii) The intersection of an arbitrary collection of closed sets is closed.

Proof. Use Theorem 3.1 and De Morgan’s Laws. 

3.2. Compact Sets

Definition 3.6. A set K ⊆ R is called compact if every sequence in K has a subsequence
that converges to a limit in K.
Theorem 3.8 (Heine-Borel Theorem (HBT)). A set K ⊆ R is compact if and only if
K is bounded and closed.

Proof. First let K be compact and we show that K is bounded and closed. Assume first,
for contradiction, K is not bounded. This means that, for every number n ∈ N, there
exists a xn ∈ K such that |xn | > n. Now, since K is compact, the sequence (xn ) in K
has a subsequence, say (xnk ), converging to a limit x ∈ K. However, since |xnk | > nk ≥ k,
this convergent subsequence is not bounded, contradicting the result that every convergent
sequence be bounded. So K must be bounded. Now we show K is closed. Assume x ∈ L(K).
Then, there exists a sequence (xn ), with xn ∈ K and xn 6= x, such that (xn ) → x. Since
K is compact, (xn ) has a convergent subsequence whose limit is in K; however, since (xn )
converges, any convergent subsequence must have the same limit as (xn ), which is x. So
x ∈ K. So L(K) ⊆ K; hence K is closed.
The proof of the converse statement is easier. For example, assume K is closed and
bounded. Let (xn ) be a sequence in K. We show that (xn ) has a subsequence converging
to some number in K. Since (xn ) is bounded, by the BW, there exists a subsequence (xnk )
converging to some number x ∈ R. Then (xnk ) is a Cauchy sequence in K. Since K is
closed, by Theorem 3.3 above, x ∈ K. Hence, by definition, K is compact. 
Example 3.2. Let K ⊆ R be compact. Show that both sup K and inf K are in K; that is,
max K and min K both exist.

Proof. Exercises! 
4 3. Basic Topology of R

Our final result is a generaliztion of (NIP).

Theorem 3.9. If K1 ⊇ K2 ⊇ K3 ⊇ · · · is a nested sequence of nonempty compact sets,
then the intersection ∩∞
n=1 Kn is nonempty.

Proof. For each n ∈ N, since Kn is nonempty, select an element xn ∈ Kn . Since xn ∈ K1

and K1 is compact, it follows that (xn ) has a subsequence (xnk ) converging to some x ∈ K1 .
We show that this x in fact belongs to every Kn for n ∈ N. Given a particular n0 ∈ N,
since nk ≥ k, we have nk ≥ n0 for all k ≥ n0 . We select a subsequence of (xnk ) consisting of
terms with k ≥ n0 ; then this subsequence also converges to x and each of its terms is also
in the compact set Kn0 . Hence the limit x ∈ Kn0 . But n0 is arbitrary; so x ∈ ∩∞ n=1 Kn and
∞
hence ∩n=1 Kn 6= ∅.