University of Illinois at Urbana-Champaign Fall 2006

Math 444 Group E13

Midterm III
Thursday, November 30.
Correction.

1. Look up your notes, or the textbook !

2. (10 points)
Are the following statements true of false ? Explain.
(a) The image of an open interval by a continuous function is an open interval.
(b) If a function f is continuous on (0, 1) then it is bounded on (0, 1).
(c) If f : R → R is continuous and takes only rational values then f is constant.
(d) There exists a dierentiable function f : R → R such that f 0 (x) = −1 for all x < 0,
f 0 (0) = 0 and f 0 (x) = 1 for all x > 0.

Correction. (a) This is false ; for instance the function dened by f (x) = 0 is continuous on
R, and the image of the open interval (0, 1) is {0}, which is not an open interval.
(b) Again, this is false ; for instance f : x 7→ x1 is continuous on (0, 1) but is not bounded
(f ( n1 ) = n for all n ∈ N and N is not bounded).
(c) This is true : indeed, if f is continuous on R then f (R) is an interval. Thus, if f takes
two values x, y with x < y , f must take all the values in the interval [x, y] ; and this interval
contains an irrational number because of the density of irrationals in the real line. Thus we
proved that if f is continuous and not constant then it takes irrational values ; this is the same
as saying that if a continuous function f takes only rational values then it is constant.
(d) By Darboux's theorem, a derivative must satisfy the conclusion of the intermediate value
theorem (i.e the image of an interval is an interval). The function here doesn't satisfy this
property (the image of R is the three-element set {−1, 0, 1}, which is not an interval). Hence
there cannot exist a function whose derivative is the function in the statement above (and
thus (d) is false).

3. (10 points)
x2 + ax + b if x ≥ 0
(
Let f be dened on R by f (x) = .
sin(x) if x < 0
Is it possible to nd a, b such that f is dierentiable on R ? If not, explain why ; if yes, give
the values of a, b.
Correction. On (−∞, 0) one has f (x) = sin(x), which is dierentiable ; on (0, +∞) one has
f (x) = x + a + b which is also dierentiable for any a, b ∈ R. Thus we only need to look at
2

what happens at 0 ; f has a limit at 0 if, and only if, its left-hand limit and its right-hand limit
at 0 are equal, thus f is continuous at 0 if, and only if, b = 0. Now assume b = 0 ; we need
to look at which condition ensures that f (x)
x
has a limit at 0. The mean value theorem shows
that if x > 0 one has x = 2c + a for some c ∈ (0, x), and if x < 0 one has f (x)
f (x)
x
= cos(c) for
some c ∈ (x, 0). Thus the right-hand limit of x at 0 is a, and its left-hand limit is at 0 is 1.
f (x)

Hence f is dierentiable at 0 if, and only if, b = 0 and a = 1. This shows that there indeed
exist a, b such that f is dierentiable on R, and that one has a = 1, b = 0.

4. (10 points) Assume that f : [0, 1] → R is a continuous function such that for all x ∈ [0, 1]
1
there exists y ∈ [0, 1] such that |f (y)| ≤ |f (x)|. Show that there exists x ∈ [0, 1] such that
2
f (x) = 0.

Correction. Let us rst solve this using general theorems about continuous functions : the
function |f | is continuous on [0, 1] since it is a composition of two continuous functions. Thus
it must have a minimum on [0, 1], i.e there must exist x such that |f (x)| ≤ |f (y)| for all
1
y ∈ [0, 1]. But we know that there must exist y such that |f (y)[≤ |f (x)| ; this implies that
2
1
|f (x)| ≤ |f (x)|, which is only possible if f (x) = 0.
2
One may also solve this exercise using the usual methods to prove results involving continuous
functions on closed bounded intervals : produce a sequence, use the Bolzano-Weierstrass theo-
rem to obtain a convergent subsequence, then use the continuity of the function to conclude.
Here, we proceed as follows : pick any x1 ∈ [0, 1], then use the property of the function to build
inductively a sequence (xn ) of elements of [0, 1] such that |f (xn+1 )| ≤ 12 |f (xn )|. By induction,
one obtains |f (xn )| ≤ 2n−1
1
|f (x1 )|. Hence (f (xn )) converges to 0. Now we can use the Bolzano-
Weierstrass theorem (the sequence (xn ) is bounded) to obtain that there exists a convergent
subsequence (xϕ(n) ) of (xn ), and its limit x must belong to [0, 1] because it is a closed interval.
But then, since f is continuous, f (xϕ(n) ) converges to f (x), which yields f (x) = 0.