Math 490 Midterm 3 (Solutions); due Friday, December 6, 2013

Problem 1.
Consider the set X = [0, 1] with the topology
T = {} {U [0, 1] | [0,
1
2
] U}
(you do not need to prove that T is a topology on X).
Determine whether or not the topological space (X, T ) is compact
and give a proof justifying your answer.
Solution.
The space (X, T ) is not compact. Indeed, for every x (
1
2
, 1] put
U
x
= [0,
1
2
] {x}. Then U
x
is open in (X, T ) and
x(
1
2
,1]
U
x
= [0, 1],
so that (U
x
)
x(
1
2
,1]
) is an open cover of X. Clearly, this cover does not
admit a nite subcover. Indeed, for any n 1, x
1
, . . . , x
n
(
1
2
, 1] we
have
n
i=1
U
x
i
= [0,
1
2
] {x
1
, . . . , x
n
} = [0, 1].
Thus (X, T ) is not compact, as claimed.
Problem 2.[10 points]
Let (X, T ) be as in Problem 1.
(a) Determine whether or not the space (X, T ) is connected and give
a proof justifying your answer.
(b) Determine whether or not there exists a metric d on X such that
the topology T is exactly the metric topology on X corresponding to
d. Give a proof justifying your answer.
Solution.
(a) Yes, (X, T ) is connected. Indeed, if U, V X are nonempty open
subsets then [0,
1
2
] U V . Therefore there do not exist nonempty
open subsets U, V X such that U V = and U V = X. Thus
X is indeed connected, as claimed.
(b) No, there does not exist a metric d on X such that that the
topology T is exactly the metric topology on X corresponding to d.
Indeed, if such a metric did exist, then (X, T ) would be a Hausdor
topological space since all metric spaces are Howsdor.
However, (X, T ) is not Hausdor. Indeed, for any open subsets U, V
of X such that 0 U and 1 V we have [0,
1
2
] U V so that
U V = .
Problem 3.
Determine whether or not the following statement is true:
1
2
If (X, d) is a bounded metric space and A, B X are nonempty
closed subsets then there exist a
0
A and b
0
B such that d(a
0
, b
0
) =
inf{d(a, b)|a A, b B}.
If the statement is true, prove it, and if the statement is false, give
a counter-example.
Solution.
The statement is false.
Let X = [1, 0) (0, 1] with d(x, y) = |x y| for x, y X.
Put A = [1, 0) and B = (0, 1]. Then A, B are closed subsets of
X since A = X (, 0] and B = X [0, ). We have d(A, B) =
inf{d(a, b)|a A, b B} = 0. However AB = and so there do not
exist a
0
A, b
0
B such that d(a
0
, b
0
) = 0.
3
Problem 4.[10 points]
(a) Give an example of a topological space X and a nonempty subset
Y X, such that Y is compact (with the subspace topology) but that
Y is not closed in X. Justify that your example has the required
properties.
(b) A metric space (X, d) is called totally bounded if for every > 0
there exist n 1, x
1
, . . . , x
n
X such that X =
n
i1
B(x
i
, ).
Give an example of a totally bounded metric space (X, d) such that
(X, d) is not compact. Justify that your example has the required
properties.
Solution.
(a) Let X = R and T = {X, } be the trivial topology on R. Then
(X, T ) is compact since any open cover contains X as one of the open
sets in this cover.
Put Y = {1}. Then Y , with the subspace topology, is compact, since
all nite topological spaces are compact. However, Y is not closed in
X since X \ Y = (, 1) (1, ) T .
(b) Let X = (0, 1] with the standard metric d(x, y) = |x y|. The
metric space (X, d) is bounded but not complete. Indeed, the sequence
(
1
n
)
n1
is a Cauchy sequence in X which does not converge in X.
The space (X, d) is totally bounded. Indeed, for any > 0 we can
choose an integer n 1 such that
1
n
< . Put x
i
=
i
n
, for i = 1, . . . , n.
Then
n
i=1
B(x
i
,
1
n
) = X and therefore
n
i1
B(x
i
, ) = X.
Problem 5.[10 points]
Let X = [0, 1] with the subspace topology T
u
coming from the upper
limit topology on R. Let T
e
be the Euclidean topology on X = [0, 1].
Give an example of a continuous function f : (X, T
u
) (X, T
e
) such
that f(0) = 0, f(1) = 1 and that there exists c (0, 1) such that
c f(X). Justify that your example has the required properties.
Solution.
Consider the function f : X X dened as
f(x) =

0 for 0 x
1
2
,
1 for
1
2
< x 1.
We claim that f : (X, T
u
) (X, T
e
) is continuous. First, observe that
the sets [0,
1
2
] and (
1
2
, 1] are open in T
u
. Indeed, the sets (,
1
2
] =

n=1
(n,
1
2
] and (
1
2
, ) =

n=1
(
1
2
+
1
n
, n] are open in the upper limit
4
topology on R. Therefore the sets [0,
1
2
] = X (,
1
2
] and (
1
2
, 1] =
X (
1
2
, ) are open in T
u
.
Let U X be an open subset in T
e
.
If 0 U and 1 U then f
1
(U) = is open in (X, T
u
).
If 0 U but 1 U then f
1
(U) = [0,
1
2
] is open in (X, T
u
).
If 0 U but 1 U then f
1
(U) = (
1
2
, 1] is open in (X, T
u
).
If 0, 1 U then f
1
(U) = X is open in (X, T
u
).
Thus for every open subset U of (X, T
e
) the set f
1
(U) is open in
(X, T
u
). Therefore f is a continuous function.
However, by construction, we have
1
100
f(X).
Problem 6.[10 points]
Let X = F(R, R) be the set of all bounded functions from R to R.
[A function f : R R is called bounded if there exists C R, C > 0
such that for every x R we have |f(x)| C.]
For f, g X dene d(f, g) := sup
xR
|f(x) g(x)|.
(a) Prove that (X, d) is a metric space.
(b) Determine whether or not the metric space (X, d) is complete.
Give a proof justifying your answer.
Solution.
(a) First, if f, g F(R, R) then there exists C > 0 such that |f(x)|
C and |g(x)| C for every x R. Therefore for every x R we have
|f(x) g(x)| |f(x)| +|g(x)| 2C
and hence d(f, g) = sup
xR
|f(x) g(x)| 2C < . Thus d is indeed
a function d : XX [0, ). By construction, for every f, g X we
have d(f, g) = d(g, f) since |f(x) g(x)| = |g(x) f(x)|.
If f X then d(f, f) = sup
xR
|f(x)f(x)| = 0. On the other hand,
if f, g X and d(f, g) = sup
xR
|f(x)g(x)| = 0 then |f(x)g(x)| = 0
for all x R, so that f(x) = g(x) for all x R and therefore f = g.
Finally, if f, g, h X, then
d(f, h) = sup
xR
|f(x) h(x)| sup
xR
(|f(x) g(x)| +|g(x) h(x)|)
sup
xR
|f(x) g(x)| + sup
xR
|g(x) h(x)| = d(f, g) +d(g, h).
Thus (X, d) is a metric space, as required.
(b) Yes, the metric space (X, d) is complete. Indeed, suppose that
(f
n
)
n1
is a Cauchy sequence in (X, d).
5
Thus for every > 0 there exists an integer N

1 such that for all

we have
(*) d(f
n
, f
m
) = sup
xR
|f
n
(x) f
m
(x)| < .
Therefore for every x R and every > 0 for all m, n N

we
have |f
n
(x) f
m
(x)| < . This means that for every x R the se-
quence (f
n
(x))
n1
is a Cauchy sequence of real numbers. Since R is a
complete metric space, for every x R the sequence (f
n
(x))
n1
has a
limit in R, denote this limit by f(x). Thus for every x R we have
lim
n
f
n
(x) = f(x) R.
We have thus constructed a function f : R R.
We claim that the function f is bounded. Indeed, for = 1 and N =
N
1
, we know that for every n N we have sup
xR
|f
n
(x) f
N
(x)| < 1.
Since f
N
X, there exists C > 0 such that for every x R we
have |f
N
(x)| C. Hence for every x R and every n N we have
|f
n
(x)| C + 1. Therefore for every x R we have
|f(x)| = | lim
n
f
n
(x)| C + 1.
Thus the function f is bounded, so that f X.
We now claim that lim
n
f
n
= f in (X, d)
Indeed, let > 0 be arbitrary and let N

1 be as in (*) above.
Fix an arbitrary x R. Then for every n, m N

we have |f
n
(x)
f
m
(x)| < . In particular, for every n N

we have f
n
(x) [f
N

(x)
, f
N

(x) + ]. Since lim

(x) , f
N

(x) + ]. Therefore for every n N

we have |f
n
(x)
f(x)| < 2.
Since x R was arbitrary, this means that for every n N

we have
d(f
n
, f) < 2. Hence, by denition of convergence in a metric space,
we have lim
n
f
n
= f in (X, d).
Thus every Cauchy sequence in (X, d) has a limit in X, so that the
metric space (X, d) is complete.