Definition: Let’s call a function “locally bounded at x”
if there exists a neighborhood of x, N(x;εx) and a number Mx
s.t. f(x) ≤ Mx in the
intersection of N(x;εx) with the domain of f.
Definition: We say that f is bounded on a set S if there is
a number M s.t. f(x) ≤ M for all x in S.
Theorem: If
f is locally bounded at each point in I, a closed interval, then f is bounded
on I.
Before we
prove this, please note that this is FALSE
for f locally bounded everywhere on an
open interval. Example?
And note
that this is false for an unbounded
interval. Example?
Proof:
Suppose f is
locally bounded on I a closed interval.
Then consider the collection of open intervals:
C = {
N(x;εx) | x in I } obtained by applying the definition of
locally bounded. Since each x in I is
the center of one of the intervals in C, then C is an “open cover” of I. But since I is compact, C contains a finite
subcover. Let N1, N2,
... Nk be the finite subset of C . I.e. each Ni is one of
the N(x;εx) from C. And
let Mi be the bounds on f(x) in Ni. Finally let M = max{Mi|i=1,...,k}. Then since each x is in some Ni , f(x)
≤ Mi ≤ M for all x in I.
QED
Comment: Note that this proof uses the finite subcover very strongly. Without finiteness we can’t choose an M as a
max.