Midterm Test suggested solution

1. Let f be a continuous function defined on [a, b] with f (a) = f (b).

(a) Suppose f 0 exists on (a, b). Show that there is ξ ∈ (a, b) such that f 0 (ξ) = 0.

Proof. As f is continuous on [a, b], f attains its maximum and minimum on [a, b].
That is to say ∃ p, q ∈ [a, b] such that

f (p) ≤ f (x) ≤ f (q) ∀ x ∈ [a, b].

If f (p) = f (q), f is a constant function. So f 0 ≡ 0.

Suppose f (p) < f (q), we may assume f (q) > f (a) = f (b). Then q is a interior
maximum. If f 0 (q) > 0, then ∃δ > 0 such that

f (q + h) − f (q)
> 0 ∀h ∈ (−δ, δ) \ {0}
h

which contradicts with the fact that f attains maximum at q. Similarly, f 0 (q)
cannot be negative. So f 0 (q) = 0.

(b) If the continuity of f at a and b is removed, does the part (i) still hold?

Proof. No. Choose f : [0, 1] → R where f (0) = f (1) = 0 and f (x) = x if

x ∈ (0, 1).

2. Let f and g be continuous functions on [a, b]. Suppose that f and g are differentiable
on (a, b) with |f 0 (x)| ≤ 1 ≤ |g 0 (x)| on (a, b). Show that |f (x) − f (a)| ≤ |g(x) − g(a)|
on [a, b].

Proof. Let x ∈ (a, b], by mean value theorem we can find c, d ∈ (a, b) such that

f (x) − f (a) = f 0 (c)(x − a) and g(x) − g(a) = g 0 (d)(x − a).

Noted that c and d may not be the same. Then

|f (x) − f (a)| = |f 0 (c)||x − a| ≤ |x − a| ≤ |g 0 (d)||x − a| = |g(x) − g(a)|.

When x = a, the inequality trivially holds.

3. Define f : [−1, 1] → R by f (t) = −1 if t < 0 and f (t) = 1 if t ≥ 0. Let

Z x
F (x) = f (t) dt
−1

for x ∈ (−1, 1]. If F differentiable on (−1, 1)?

1
 R a+b
Proof. f is clearly integrable. So F (a + b) = F (a) + a
f for a, b and a + b ∈ [−1, 1].
For h ∈ (−1, 1) ,
h
F (h) − F (0)
Z
1
= · f (t) dt.
h h 0

If h > 0, then
h
F (h) − F (0)
Z
1
= · 1 dt = 1.
h h 0

If h < 0,
h
F (h) − F (0)
Z
1
= · −1 dt = −1.
h h 0

So F is not differentiable at x = 0. By fundamental theorem of Calculus, F is

differentiable on c where f is continuous. So F is differentiable on (−1, 1) \ {0}.
√
4. If f is nonnegative Riemann integrable function on [a, b], does it imply f Riemann
integrable on [a, b]?

Proof. Yes. Since f ∈ R[a, b], for any  > 0, there exists δ > 0 such that whenever P
is a partition with ||P|| < δ,

U (f, P) − L(f, P) < 2 .

For such P, denote (mi )Mi = (inf) sup{f (x) : x ∈ [xi , xi+1 ]}.
p √ √
And also (m̃i )M̃i = (inf) sup{ f (x) : x ∈ [xi , xi+1 ]} = ( mi ) M i .
p p n
X
U ( f , P) − L( f , P) = (M̃i − m̃i )∆xi
i=1
n
X √ √
= ( M i − mi )∆xi .
i=1

We split the sum into two parts.

 
p p X X √ √
U ( f , P) − L( f , P) =  +  ( M i − mi )∆xi
Mi ≥2 Mi <2

1 X X √ √
≤ (Mi − mi )∆xi + ( M i − mi )∆xi
 2 2
Mi ≥ Mi <

n n
1 X X
≤ · (Mi − mi )∆xi +  · ∆xi
 i=1 i=1

≤  +  · (b − a) = (b − a + 1).

To conclude, ∀  > 0, ∃ partition P on [a, b] such that

p p
U ( f , P) − L( f , P) < (b − a + 1).

2
 R1
5. Suppose f is Riemann integrable on [0, 1], find limn→∞ 0
xn f (x) dx

Proof. The limit is zero. Since f is Riemann integrable, f is bounded. Let M > 0
such that |f (x)| ≤ M on [0, 1].
Z 1 Z 1
n M
x f (x) dx ≤ M xn dx = → 0.
0 0 n+1