Sol HMK9
Sol HMK9
1
#1) By the definition of continuity, f is continuous at c if and only if limxc f = f (c). By the Sequential
Criterion in Chapter 4, this is equivalent to the statement that if (xn ) is a sequence converging to c
such that xn 6= c, then (f (xn )) converges to f (c). Clearly, this condition is satisfied if the Sequential
condition in 5.1.3 (which allows for ANY sequence converging to c) is satisfied. Conversely, let (xn ) be
any sequence converging to c. Either it is ultimately constant and equal to c, and there is nothing to
prove since f (xn ) will be ultimately constantly f (c), or the terms xnk not equal to c form an infinite
subsequece converging to c. Thus, if f is continuous at c, f (xnk ) converges to f (c), and hence f (xn )
converges to f (c), since every term not in f (xnk ) is equal to f (c).
#3) Let > 0. If x0 [a, b), then we can find > 0 such that x0 + < b, and if x [a, b], |x x0 | < ,
then |f (x) f (x0 )| < . (Note that if x [a, c], |x x0 | < , then automatically, x [a, b).) Since
h f on [a, b), then for x [a, c], if |x x0 | < , then |h(x) h(x0 )| = |f (x) f (x0 )| < . So h is
continuous at every point in [a, b).
Similarly, h is continuous at every point in (b, c].
To show that h is continuous at b, note that we can find 1 , 2 > 0 such that if x1 [a, b], x2 [b, c],
and |xi b| < i for i = 1, 2, then |f (x1 ) f (b)|, |g(x2 ) g(b)| < . Choosing = min{1 , 2 }, and
noting that |f (x1 ) f (b)| = |h(x1 ) h(b)| and |g(x2 ) g(b)| = |h(x1 ) h(b)|, we conclude that if
x [a, c] and |x b| < , then |h(x) h(b)| < .
2
+x6
= x + 3. So limx=2 f = 5. Therefore, by defining f (2) = 5, we
#5) Away from x = 2, we see that x x2
extend f to a continuous function on the real line.
#10) Let > 0. Let > 0. Note that ||x| |c|| |x c|, so if |x c| < , then ||x| |c|| < , So |x| is
continuous at every point c.
#15) Since the limx0 f does not exist, there is some sequence an > 0 such that an converges to 0, but
f (an ) diverges. Since f is a bounded function, f (an ) is a bounded sequence. By Bolzano-Weierstrass,
we can find a convergent subsequence f (ank ), which corresponds to a convergent subsequence ank . By
3.4.9, since f (an ) is a bounded subsequence, if all of its convergent subsequences had the same limit,
then f (an ) would be convergent, contradicting the fact that we chose it to be divergent. Therefore, we
can find at least two subsequences f (ank ) and f (anj ) that converge to different limits. Let xk = ank
and yj = anj .
Section 5.2
#1) All 4 functions are continuous wherever they are defined.
1
x>c
#3) Define f (x) =
And define g(x) = f (x). Then f and g are both discontinuous at c,
1 x c
f + g is constant 0 and f g is constant 1, which are both continuous.
#6) Let (xn ) be an arbitrary sequence converging to c that is not ever equal to c. By the Sequential
Criterion of Convergence, f (xn ) converges to b. By the Sequential Criterion for Continuity applied to
this sequence and g, g(f (xn )) converges to g(b). Therefore, by the sequential criterion for convergence,
g f (x) g(b) as x c.
#7) Let f (x) = 1 if x is rational, and f (x) = 1 if x is irrational. Then f is discontinuous at every point,
but |f | is the constant function 1.
#12) Say f is additive and continuous at x0 , and let c be any other point. As x c, x c + x0 x0 .
Since f is continuous at x0 , this implies that limxc f (x) = limxc [f (x c + x0 ) + f (c x0 )] =
f (x0 ) + f (c x0 ) = f (c), using additivity twice. Thus, f is continuous at c.
#13) First, note that f (0) = 0, since f (0) = f (0 + 0) = f (0) + f (0) = 2f (0). Also note that 0 = f (0) =
f (x + x) = f (x) + f (x) so that f (x) = f (x). This last comment shows that it suffices to prove
the claim for positive numbers, and it will follow automatically for negative numbers.
Let c = f (1). Let m N. Then c = f (1) = f (1/m + 1/m + + 1/m) = f (1/m) + f (1/m) + +
f (1/m) = mf (1/m), where the each some has m terms. Thus, f (1/m) = c/m. Now let n N. Then
f (n/m) = f (1/m + + 1/m) = f (1/m) + + f (1/m) = nf (1/m) = c(n/m), where now each sum
has n terms. Therefore, f (q) = cq for all rational numbers q.
Let x R be arbitrary. Let qn be a sequence of rationals converging to x (this exists by the Density
Theorem). Since qn is rational, f (qn ) = cqn . By The continuity of f , we pass to the limit on both
sides to find that f (x) = cx as desired.