Lecture 01 Functional Analysis Week01
Lecture 01 Functional Analysis Week01
(X \ U
), (1)
X \
(X \ U
). (2)
Problem 1 Prove relations (1) and (2).
According to de Morgans laws, it follows from Denition 1 that:
1. The space (X, T) itself and the empty set are closed;
2. Arbitrary (nite or innite) intersections
is also a
topology in X.
Proof. We need to verify Denition 1 for T. Clearly
is stable by the operations of taking arbitrary unions and nite intersections, the same is true for
.
Corollary 1 Let A be any system of subsets of a set X. Then there exists a minimal topology in X
containing A, i.e., a topology T(A) containing A and contained in every topology containing A.
Proof. A topology containing A always exists, e.g., the discrete topology in which every subset of
X is open. The intersection of all topologies containing A is the desired minimal topology T(A),
often called the topology generated by the system A.
Denition 11 Let A be a system of subsets of X and A a xed subset of X. Then the system A
A
consisting of all subsets of X of the form A B, B A is called the trace of the system A on the
set A.
It is easy to see that the trace (on A) of a topology T (dened in X) is a topology T
A
in A. (Such
a topology is often called a relative or induced topology.) In this sense, every subset A of a given
Week 1: Topological spaces 5
topological space (X, T) generates a new topological space (A, T
A
), called a subspace of the original
topological space (X, T).
Example 8 The set ]0, 1] R is not open in the usual topology of R, but it is open in the induced
topology when considered as a subset of A = [1, 1].
Let us also notice that if T
1
and T
2
are dierent topologies in X, they can generate the same relative
topology T
A
in A.
Denition 12 Let (X, T
X
) and (Y, T
Y
) be two topological spaces. The product topology in XY ,
denoted T = T
X
T
Y
, is dened by calling U X Y open if
(x, y) U V T
X
and W T
Y
such that x V, y W and V W U.
Example 9 R
2
= R R is a topological space with the usual product topology constructed on the
two usual topologies in R.
2 Lecture 1.4: Convergence and continuity
2.1 Continuous mappings. Homeomorphism
Denition 13 Let f be a mapping of one topological space (X, T
X
) into another topological space
(Y, T
Y
), so that f associates an element y = f(x) Y with each element x X. Then f is said to
be continuous at the point x
0
X if, given any neighborhood W
y
0
of the point y
0
= f(x
0
), there
is a neighborhood U
x
0
of the point x
0
such that f(U
x
0
) W
y
0
.
Denition 14 The mapping f : (X, T
X
) (Y, T
Y
) is said to be continuous on X if it is contin-
uous at every point of X.
In particular, a continuous mapping of a topological space (X, T
X
) into the real line (R, T
R
) is called
a continuous real mapping on X. Here by T
R
we denote the usual topology on R (see Denition 4).
The notion of continuity of a mapping f of one topological space into another is easily stated in
terms of open sets, i.e., in terms of the topologies of the two spaces:
Theorem 3 A mapping f of a topological space (X, T
X
) into a topological space (Y, T
Y
) is continuous
if and only if the inverse image f
1
(W) of every open set W T
Y
is open, i.e. f
1
(W) T
X
.
Proof. Suppose f is continuous on (X, T
X
), and let W be any open subset of T
Y
. Choose any point
x f
1
(W), and let y = f(x) (see Fig. 1). Then W is an open neighborhood of the point y. Hence,
by the continuity of f, there is a neighborhood U
X
of x such that f(U
X
) W, i.e., U
X
f
1
(W).
In other words, every point x f
1
(W) has a neighborhood contained in f
1
(W). Consequently
f
1
(W) is open.
Conversely, suppose f
1
(W) is open whenever W Y is open. Given any point x X, let U
Y
be
any open neighborhood of the point y = f(x).
Then clearly x f
1
(U
Y
), and moreover f
1
(U
Y
) is open, by hypothesis. Therefore U
X
= f
1
(U
Y
)
is a neighborhood of x such that f(U
X
) U
Y
. In other words, f is continuous at x and hence on X,
since x is an arbitrary point of X.
Let us recall the following Lemma:
6 Week 1: Topological spaces
U
x
x
y
W
f
f
1
f(U
x
)
f
1
(W)
Figure 1 Illustration of the proof of Theorem 3: for any open W in Y , any point x f
1
(W) is associated
with y = f(x); U
X
is a neighborhood of x such that f(U
X
) W.
Lemme 1 1. The inverse image of a union (or intersection) of sets equals the union (or inter-
section) of the inverse images of the sets:
f
1
(A B) = f
1
(A) f
1
(B), f
1
(A B) = f
1
(A) f
1
(B) A, B Y.
2. The inverse image of the complement of a set is the complement of the inverse image of the
set:
U Y f
1
(Y \ U) = X \ f
1
(U).
Remark 2 Suppose f : (X, T
X
) (Y, T
Y
) is a mapping of a topological space (X, T
X
) into a
topological space (Y, T
Y
). Let f
1
(T
Y
) be the inverse image of the topology T
Y
:
f
1
(T
Y
) = { system of all sets f
1
(U) | U T
Y
}.
Thanks to the point 1 of Lemma 1, we obtain that f
1
(T
Y
) is a topology in X.
Problem 4 Prove the following theorem:
Theorem 4 Let f : (X, T
X
) (Y, T
Y
) be a mapping of a topological space (X, T
X
) into a topological
space (Y, T
Y
). The mapping f is continuous if and only if the topology T
X
is stronger than the topology
f
1
(T
Y
).
Thanks to point 2 of Lemma 1, we obtain the dual form of Theorem 3:
Theorem 5 A mapping f of a topological space (X, T
X
) into a topological space (Y, T
Y
) is continuous
if and only if the inverse image f
1
(W) of every closed set W Y is closed in X.
It is important to notice that the image (as opposed to the inverse image) of an open set under
a continuous mapping need not be open. Similarly, the image of a closed set under a continuous
mapping need not be closed. As a direct Corollary of Theorem 3, we have the theorem on continuity
of composite mapping:
Theorem 6 Given topological spaces (X, T
X
), (Y, T
Y
) and (Z, T
Z
), suppose f is a continuous map-
ping of (X, T
X
) into (Y, T
Y
) and g is a continuous mapping of (Y, T
Y
) into (Z, T
Z
). Then the mapping
g f: x X g(f(x)) Z is continuous.
Denition 15 Given two topological spaces (X, T
X
) and (Y, T
Y
), let f be a bijection of (X, T
X
) onto
(Y, T
Y
), and suppose f and f
1
are both continuous. Then f is called a homeomorphic mapping
or simply a homeomorphism (between X and Y ). Two spaces (X, T
X
) and (Y, T
Y
) are said to be
homeomorphic if there exists a homeomorphism between them. We note the homeomorphic spaces by
(X, T
X
) (Y, T
Y
).
Week 1: Topological spaces 7
Homeomorphic spaces have the same topological properties, and from the topological point of view
are merely two representatives of one and the same space. In fact, if f is a homeomorphic mapping
of (X, T
X
) onto (Y, T
Y
), then T
X
= f
1
(T
Y
) and T
Y
= f(T
X
).
Example 10 The interval ]
2
,
2
[ equipped with the usual topology on R is homeomorphic to R also
equipped with the usual topology, as f(x) = tan(x) is a homeomorphic mapping of ]
2
,
2
[ into R.
Remark 3 The relation of being homeomorphic is reexive, i.e. for any topological space
(X, T
X
),
(X, T
X
) (X, T
X
),
symmetric, i.e. for all topological spaces (X, T
X
) and (Y, T
Y
),
(X, T
X
) (Y, T
Y
) (Y, T
Y
) (X, T
X
),
and transitive, i.e. for all topological spaces (X, T
X
), (Y, T
Y
) and (Z, T
Z
),
(X, T
X
) (Y, T
Y
) and (Y, T
Y
) (Z, T
Z
) (X, T
X
) (Z, T
Z
),
and hence is called an equivalence relation. Therefore any given family of topological spaces can
be partitioned into disjoint classes of homeomorphic spaces.
2.2 Converging sequences in (X, T)
Let V(x) denote the set of all neighborhoods of x X and O(x) denote the set of all open neigh-
borhoods of x X.
Denition 16 Let (X, T) be a topological space and (x
n
) be a sequence of elements of X. We say
that (x
n
) converges to l, if
U V(l), N N such that n N x
n
U.
We note that
(x
n
) may converge to several elements of X;
If the topology on X is stronger (larger/ner), it is harder for (x
n
) to converge;
If X is equipped with the discrete topology, only sequences that become constant converge.
Remark 4 Thanks to the denition of the neighborhood U, as any set containing an open set V
such that x V (and thus V is an open neighborhood of x!), we can consider the equivalent form of
Denition 16 choosing only open neighborhoods of l:
Denition 17 Let (X, T) be a topological space and (x
n
) be a sequence of elements of X. We say
that (x
n
) converges to l if
V O(l), N N such that n N x
n
V.
For the sake of clarity, in what follows we will use Denition 17.
Proposition 1 Let (X, T
X
) and (Y, T
Y
) be two topological spaces. Let f : X Y be a continuous
mapping and (x
n
) be a sequence in X converging to l. Dene y
n
= f(x
n
). Then (y
n
) converges to
f(l) in Y .
8 Week 1: Topological spaces
Proof. Let U be any open neighborhood of f(l). Since f is continuous, f
1
(U) is open (thanks to
the denition of continuity). Since f(l) U, we have l f
1
(U) and consequently f
1
(U) is a open
neighborhood of l. Since (x
n
) converges to l in X,
N N such that n N x
n
f
1
(U),
implies y
n
= f(x
n
) U for all n N. As U is an arbitrary chosen open neighborhood of f(l), we
nd that
U O(f(l)), N N such that n N y
n
U.
So (y
n
) converges to f(l).
Example 11 Let us consider the topological space (X, T
1
) from Example 1. We denote by [x] the
integer part of x. The sequence (x
n
)
n1
is dened as
x
n
=
4
n
+
(1)
n
+ 3
2
,
from where we nd
x
1
= 5, x
2
= 4, x
3
= 2, x
4
= 3, x
5
= 1, x
6
= 2, x
7
= 1, x
8
= 2, . . .
We notice that for n 2 x
2n+1
= 1 and for n 3 x
2n
= 2. It means that for n 5 x
n
{1, 2}. The
set {1, 2} is the smallest open neighborhood containing 1 or 2 in (X, T
1
):
1, 2 {1, 2} {1, 2, 3, 4} X.
Therefore, (x
n
) converge to 1 and to 2 (the limit l is not unique) in (X, T
1
).
Example 12 Let us consider R with the usual topology. The sequence (
1
n
)
n1
converge to 0 in R:
for any open neighborhood of zero ]
1
m
,
1
m
[ with m R \ {0} there exists n
0
N such that for all
n > n
0
we have
1
n
]
1
m
,
1
m
[.
3 Lecture 1.5: Final topology
We can equip a set X with a topology that makes every mapping f
i
on X continuous. If all else fails,
the discrete topology will work!
Denition 18 Let f
i
: X Y
i
be given mappings dened on a set X (i I). We call nal
topology, noted (X, {f
i
, i I}), the coarsest topology in X that makes every f
i
continuous, as
mappings from the topological space (X, (X, {f
i
, i I})) to the topological space (Y
i
, T
Y
i
).
Example 13 Let X = R, Y = R and f be dened by
f(x) =
0 if x 0
1 if x > 0
.
Equip Y with the usual topology. The nal topology on X for f is
(R, {f}) = {, ] , 0[, ]0, +[, R}.
Week 1: Topological spaces 9
Lemme 2 Let f
i
: X Y
i
be given mappings dened on a set X (i I nite or not) with image in
sets Y
i
equipped with topologies T
Y
i
. Then for any open set U = of the nal topology (X, {f
i
, i
I}) there exists a nite number of open sets V
i
in T
Y
i
such that
iG
f
1
i
(V
i
) U,
where G is a nite part of I.
Proof. Since U is open, then U (X, {f
i
, i I}) and consequently, by denition of the nal
topology, U can be presented as a nite intersection or a union of the inverse images of open sets in
Y
i
. In other words, for all U (X, {f
i
, i I}) there exist open sets V
i
in T
Y
i
such that
U =
iG
f
1
i
(V
i
) or U =
iJ
f
1
i
(V
i
),
where by G is denoted a nite part of I and by J is denoted any subset nite or not of I. Since
iJ
f
1
i
(V
i
)
iG
f
1
i
(V
i
) for a nite index set G J,
then we can always say that for all open U there exists a nite number of open sets V
i
in T
Y
i
such
that
U
iG
f
1
i
(V
i
).
Remark 5 Let f
i
: X Y
i
. We notice that the nal topology (X, {f
i
, i I}) is constructed on the
family of sets f
1
i
(V
i
) for open V
i
in Y
i
(see Section 1.1). Moreover, the family of sets f
1
i
(V
i
) is a
base of the nal topology (see H. Brezis Functional Analysis, Sobolev Spaces and Partial Dierential
Equations p. 56 or A.N. Kolmogorov, S.V. Fomin Introductory Real Analysis p.80).
Proposition 2 Let f
i
: X Y
i
be given mappings dened on a set X (i I nite or not) with
image in sets Y
i
equipped with topologies T
Y
i
and let (x
n
) be a sequence of X. In the topological space
(X, (X, {f
i
, i I})), equipped with the nal topology, for n
x
n
x if and only if i I, f
i
(x
n
) f
i
(x) [in the topological spaces (Y
i
, T
Y
i
)].
Proof. The direct statement follows from Proposition 1 as all functions f
i
are continuous for the
nal topology (X, {f
i
, i I}).
Let us prove the converse. Let U be an open neighborhood of x in (X, (X, {f
i
, i I})) and V
i
be
open neighborhoods of f
i
(x) in (Y
i
, T
Y
i
). As mappings f
i
are continuous for all i I, the inverse
images f
1
i
(V
i
) are open thanks to Theorem 3. In addition, x f
1
i
(V
i
) for all i I, thus the inverse
images f
1
i
(V
i
) and also any nite intersection of them are open neighborhoods of x. Therefore,
using Lemma 2, there exists a nit subset G of I such that the nite intersection of
iG
f
1
i
(V
i
) is
contained in U.
As f
i
(x
n
) f
i
(x) for all i, then
there exists N
i
N such that n N
i
f
i
(x
n
) V
i
.
Let N be the largest N
i
for all i G. Then n N implies x
n
iG
f
1
i
(V
i
) U, which completes
the proof.
10 Week 1: Topological spaces
4 Lecture 1.6: What a topology sees and does not see.
Separation of topological spaces
Let (X, T
X
) be a topological space. We would like to address these questions:
If x is a limit point of A X. Is there a sequence (x
n
) of points of A converging to x?
Let x
n
x in (X, T
X
). IS the limit unique?
Let x and y be two dierent points of (X, T
X
). Can we separate them?
To do so, we need to specify the separation properties of the topological space (X, T
X
).
x
y
O
x
O
y
Figure 2 First axiom of separation.
Denition 19 A topological space (X, T
X
) said to satisfy the First axiom of separation (or to
be T
1
space) if for all two distinct points x and y in (X, T
X
) there exists an open neighborhood O
x
of the point x such that y / O
x
and there exists a open neighborhood O
y
of the point y such that
x / O
y
. (see Fig.2)
Example 14 The topological space (X, T
X
) constructed in Example 5 is not a T
1
-space.
In a T
1
-space singleton point is a closed set. Indeed, if x = y, then there exists an open neighborhood
O
y
of the point y such that x / O
y
, i.e. y = x. Thus, x = x. Consequently, in a T
1
-space any nite
union of points is a closed set.
In the topological spaces which are not T
1
-spaces, even sets composed only of a nite number of
points can possess limit points. In the topological space (X, T
X
) constructed in Example 5, the point
{a} is the limit point for the set W = {b}.
But in a T
1
-space, it holds
Lemme 3 Point x is a limit point of the set W in a T
1
-space if and only if all open neighborhoods
U of x contains innite number of points of W.
Proof. If any open neighborhood U of x contains an innite number of points from W it is obvious
(see Denition 6) that in this case x is a limit point of W. Let us prove the converse. Let x be a limit
point of W. Suppose that there exists an open neighborhood U of x such that U contains only a nite
number of points {x
1
, . . . , x
n
} of W (where x
i
= x for all i in the case when x W). As in a T
1
-space
any nite union of points is a closed set, {x
1
, . . . , x
n
} is closed. Therefore, O = U \ {x
1
, . . . , x
n
} is
open and, thus, O is an open neighborhood of x such that O(W \ {x}) = , which contradicts the
denition of a limit point.
Let us introduce Hausdor spaces.
Week 1: Topological spaces 11
Denition 20 A topological space (X, T
X
) is a Hausdor space (or a T
2
-space or a separated
space) if all two distinct points in X have two disjoint neighborhoods. (see Fig.3)
x
y
O
x
O
y
Figure 3 Hausdor space.
The main advantage of a Hausdor space is that the limit of a sequence is unique.
All Hausdor spaces are T
1
-space, but not converse.
Example 15 (T
1
-space, but not T
2
-space) Let us consider the interval [0, 1] and all sets
A
m,a
1
,...,am
= [0, 1] \ (
m
i=1
{a
i
}), where a
i
[0, 1] and 0 m +.
If X = [0, 1] with the topology T composed by and all sets A
m,a
1
,...,am
, then (X, T) is T
1
-space, but
not a Hausdor space.
x
M
O
x
O
M
Figure 4 Third axiom of separation
Denition 21 A topological space (X, T
X
) is said to satisfy the Third axiom of separation (or
to be T
3
space) if for all points x and closed sets M in (X, T
X
) not containing x, there exist two
disjoint neighborhoods O
x
of the point x and O
M
of the set M. (see Fig.4)
We note that the open neighborhood of a set M in the topological space (X, T) is called any open
set U containing M.
Denition 22 Topological spaces which satisfy the axioms T
1
and T
3
are called regular.
Obviously, each regular space is a Hausdor space. But not converse:
Example 16 (Hausdor spaces which are not regular) Let X = [0, 1]. Let all points of X
dierent to 0 have the usual neighborhoods of the usual topology. Dene the neighborhoods of zero as
all semi-intervals [0, [ without points
1
n
, n N
} such
that A
.
2. A cover {U
} of A
(b) {V
} is a cover of A.
Denition 25 A topological space (X, T) is said to be compact if every open cover of (X, T) has
a nite subcover.
Example 17 Any closed bounded subset of R
n
is compact (it will be proved later). On the other hand,
R
n
itself (e.g., the real line, a two-dimensional plane or three-dimensional space) is not compact.
Denition 26 A system of subsets {U
= X \ F
are open.
Since any nite intersection
n
i=1
F
i
is not empty, de Morgans law (2) implies
n
i=1
U
i
=
n
i=1
X \ F
i
= X \
n
i=1
F
i
X
that there is no nite system of sets U
i
= X \ F
i
which covers X. But X is compact, thus the whole
system of {U
= .
Conversely, suppose every centered system of closed subsets of (X, T) has a non-empty intersection,
and let {U
= X \ U
(X \ U
) = X \ (
) = X \ X = .
By the hypothesis, this implies that the system{F
} be any centered system of closed subsets of the subspace F X by the induced topology.
Then every F
nN
J
n
= ,
i.e., (X, T) is not compact.
Denition 27 A subset K X of a topological space (X, T) is called sequential ly compact if
every sequence in K has a convergent subsequence, or equivalently, if every innite subset of K has
at least one limit point (in K).
Thus Theorem 11 says that every compact set is sequentially compact. The converse, however, is not
true. For the relation between the concepts of compactness and sequentially compactness see A.N.
Kolmogorov, S.V. Fomin Introductory Real Analysis. We just formulate the following Theorem
(see Kolmogorov, Fomin p.95 for the proof):
Theorem 12 A topological space (X, T) is sequentially compact
1. if and only if every countable open cover of X has a nite subcover.
2. if and only if every countable centered system of closed subsets of X has a nonempty intersec-
tion.
5.2 Continuous mappings of compact spaces
Next we show that the "continuous image" of a compact space is itself a compact space:
Theorem 13 Let (X, T
X
) be a compact space and f a continuous mapping of (X, T
X
) in a topological
space (Y, T
Y
). Then f(X) endowed with the induced topology T
y
f(X) is itself compact.
Week 1: Topological spaces 15
Proof. Let {V
= f(X),
and let U
= f
1
(V
). As f is continuous, U
) =
f
1
(V
) =
.
Since (X, T
X
) is compact, {U