28.04.2014 Views

The Riemann Integral

The Riemann Integral

The Riemann Integral

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

10 1. <strong>The</strong> <strong>Riemann</strong> <strong>Integral</strong><br />

Proof. If P = {I 1 ,I 2 ,...,I n } is a partition of [a,b], then<br />

n∑<br />

[ ]<br />

U (f;P)−L(f;P) = supf −inff<br />

·|I k |<br />

I k<br />

I k<br />

=<br />

≤ C<br />

k=1<br />

n∑<br />

k=1<br />

osc<br />

I k<br />

f ·|I k |<br />

n∑<br />

k=1<br />

osc<br />

I k<br />

g ·|I k |<br />

≤ C[U(g;P)−L(g;P)].<br />

Thus, f satisfies the Cauchy criterion in <strong>The</strong>orem 1.14 if g does, which proves that<br />

f is integrable if g is integrable.<br />

□<br />

We can also give a sequential characterization of integrability.<br />

<strong>The</strong>orem 1.17. A bounded function f : [a,b] → R is <strong>Riemann</strong> integrable if and<br />

only if there is a sequence (P n ) of partitions such that<br />

In that case,<br />

∫ b<br />

a<br />

lim [U(f;P n)−L(f;P n )] = 0.<br />

n→∞<br />

f = lim<br />

n→∞ U(f;P n) = lim<br />

n→∞ L(f;P n).<br />

Proof. First, suppose that the condition holds. <strong>The</strong>n, given ǫ > 0, there is an<br />

n ∈ N such that U(f;P n ) − L(f;P n ) < ǫ, so <strong>The</strong>orem 1.14 implies that f is<br />

integrable and U(f) = L(f).<br />

Furthermore, since U(f) ≤ U(f;P n ) and L(f;P n ) ≤ L(f), we have<br />

0 ≤ U(f;P n )−U(f) = U(f;P n )−L(f) ≤ U(f;P n )−L(f;P n ).<br />

Since the limit of the right-hand side is zero, the ‘squeeze’ theorem implies that<br />

It also follows that<br />

lim U(f;P n) = U(f) =<br />

n→∞<br />

lim L(f;P n) = lim U(f;P n)− lim [U(f;P n)−L(f;P n )] =<br />

n→∞ n→∞ n→∞<br />

Conversely, if f is integrable then, by <strong>The</strong>orem 1.14, for every n ∈ N there<br />

exists a partition P n such that<br />

∫ b<br />

0 ≤ U(f;P n )−L(f;P n ) < 1 n ,<br />

a<br />

f<br />

∫ b<br />

a<br />

f.<br />

and U(f;P n )−L(f;P n ) → 0 as n → ∞.<br />

□<br />

Note that if the limits of U(f;P n ) and L(f;P n ) both exist and are equal, then<br />

lim [U(f;P n)−L(f;P n )] = lim U(f;P n)− lim L(f;P n),<br />

n→∞ n→∞ n→∞<br />

so the conditions of the theorem are satisfied. Conversely, the proof of the theorem<br />

shows that if the limit of U(f;P n ) −L(f;P n ) is zero, then the limits of U(f;P n )

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!