Académique Documents
Professionnel Documents
Culture Documents
Series
1. What is a Series?
The idea behind adding up an infinite collection of numbers is a reduction to
the well-understood idea of a sequence. This is a typical approach in mathemat-
ics: reduce a question to a previously solved problem.
D EFINITION 4.1. Given a sequence a n , the series having a n as its terms is the
new sequence
n
X
sn = ak = a1 + a2 + + an .
k=1
The numbers s n are called the partial sums of the series. If s n ! S 2 R, then the
series converges to S. This is normally written as
1
X
a k = S.
k=1
4-1
4-2 Series
called a geometric series. The number r is called the ratio of the series.
Suppose a n = r n1 for r 6= 1. Then,
s 1 = 1, s 2 = 1 + r, s 3 = 1 + r + r 2 , . . .
In general, it can be shown by induction (or even long division of polynomials)
that
Xn Xn 1rn
(4.1) sn = ak = r k1 = .
k=1 k=1 1r
The convergence of s n in (4.1) depends on the value of r . Letting n ! 1, its
apparent that s n diverges when |r | > 1 and converges to 1/(1 r ) when |r | < 1.
When r = 1, s n = n ! 1. When r = 1, its essentially the same as Example 4.1,
and therefore diverges. In summary,
1
X c
c r n1 =
n=1 1r
for |r | < 1, and diverges when |r | 1. This is called a geometric series with ratio r .
steps
2 1 1/2 0
distance from wall
In some cases, the geometric series has an intuitively plausible limit. If you
start two meters away from a wall and keep stepping halfway to the wall, no
number of steps will get you to the wall, but a large number of steps will get you
as close to the wall as you want. (See Figure 4.1.) So, the total distance stepped
has limiting value 2. The total distance after n steps is the nth partial sum of a
geometric series with ratio r = 1/2 and c = 1.
P
E XAMPLE 4.3 (Harmonic Series). The series 1 n=1 1/n is called the harmonic
series. It was shown in Example 3.18 that the harmonic series diverges.
Many have made the mistake of reading too much into Corollary 4.3. It can
only be used to show divergence. When the terms of a series do tend to zero, that
does not guarantee convergence. Example 4.3, shows Theorem 4.2(c) is necessary,
but not sufficient for convergence.
Another useful observation is that the partial sums of a convergent series are
a Cauchy sequence. The Cauchy criterion for sequences can be rephrased for
series as the following theorem, the proof of which is Exercise 4.4.
P
T HEOREM 4.4 (Cauchy Criterion for Series). Let a n be a series. The following
statements are equivalent.
P
(a) a n converges.
(b) For every " > 0 there is an N 2 N such that whenever n m N , then
X n
a i < ".
i =m
2. Positive Series
Most of the time, it is very hard or impossible to determine the exact limit of
a convergent series. We must satisfy ourselves with determining whether a series
converges, and then approximating its sum. For this reason, the study of series
usually involves learning a collection of theorems that might answer whether a
given series converges, but dont tell us to what it converges. These theorems are
usually called the convergence tests. The reader probably remembers a battery
of such tests from her calculus course. There is a myriad of such tests, and the
standard ones are presented in the next few sections, along with a few of those
less widely used.
Since convergence of a series is determined by convergence of the sequence
of its partial sums, the easiest series to study are those with well-behaved partial
sums. Series with monotone sequences of partial sums are certainly the simplest
such series.
P
D EFINITION 4.5. The series a n is a positive series, if a n 0 for all n.
2.1. The Most Common Convergence Tests. All beginning calculus courses
contain several simple tests to determine whether positive series converge. Most
of them are presented below.
2.1.1. Comparison Tests. The most basic convergence tests are the compari-
son tests. In these tests, the behavior of one series is inferred from that of another
series. Although theyre easy to use, there is one often fatal catch: in order to use
a comparison test, you must have a known series to which you can compare the
mystery series. For this reason, a wise mathematician collects example series
for her toolbox. The more samples in the toolbox, the more powerful are the
comparison tests.
P P
T HEOREM 4.6 (Comparison Test). Suppose a n and b n are positive series
with a n b n for all n.
P P
(a) If b n converges, then so does a n .
P P
(b) If a n diverges, then so does b n .
a 1 + a 2 + a 3 + a 4 + a 5 + a 6 + a 7 + a 8 + a 9 + + a 15 +a 16 +
| {z } | {z } | {z }
2a 2 4a 4 8a 8
a 1 + a 2 + a 3 + a 4 + a 5 + a 6 + a 7 + a 8 + a 9 + + a 15 +a 16 +
|{z} | {z } | {z } | {z }
a 2 2a 4 4a 8 8a 16
F IGURE 4.2. This diagram shows the groupings used in inequality (4.3).
P P
P ROOF. Let A n and B n be the partial sums of a n and b n , respectively. It
follows from the assumptions that A n and B n are increasing and for all n 2 N,
(4.2) An Bn .
P P
If b n = B , then (4.2) implies B is an upper bound for A n , and a n con-
verges.
P
On the other hand, if a n diverges, A n ! 1 and the Sandwich Theorem
3.9(b) shows B n ! 1.
P
E XAMPLE 4.5. Example 4.3 shows that 1/n diverges. If p 1, then 1/n p
P
1/n, and Theorem 4.6 implies 1/n p diverges.
P
E XAMPLE 4.6. The series sin2 n/2n converges because
sin2 n 1
n
2n 2
P
for all n and the geometric series 1/2n = 1.
1The series P 2n a P
2n is sometimes called the condensed series associated with an .
P
E XAMPLE 4.7 (p-series). For fixed p 2 R, the series 1/n p is called a p-series.
The special case when p = 1 is the harmonic series. Notice
X 2n X 1p n
= 2
(2n )p
is a geometric series with ratio 21p , so it converges only when 21p < 1. Since
21p < 1 only when p > 1, it follows from the Cauchy Condensation Test that
the p-series converges when p > 1 and diverges when p 1. (Of course, the
divergence half of this was already known from Example 4.5.)
The p-series are often useful for the Comparison Test, and also occur in many
areas of advanced mathematics such as harmonic analysis and number theory.
P P
T HEOREM 4.8 (Limit Comparison Test). Suppose a n and b n are positive
series with
an an
(4.5) = lim inf lim sup = .
bn bn
P P P
(a) If 2 (0, 1) and a n converges, then so does b n , and if b n di-
P
verges, then so does a n .
P P P
(b) If 2 (0, 1) and b n diverges, then so does a n , and if a n con-
P
verges, then so does b n .
Then
an 1/(2n n) 2n 1
lim= lim n
= lim n
= lim = 1 2 (0, 1).
n!1 b n n!1 1/2 n!1 2 n n!1 1 n/2n
P
Since 1/2n = 1, the original series converges by the Limit Comparison Test.
P ROOF. First, suppose < 1 and r 2 (, 1). There is an N 2 N so that a n1/n < r
for all n N . This is the same as a n < r n for all n N . Using this, it follows that
when n N ,
n
X NX
1 n
X NX
1 n
X NX
1 rN
ak = ak + ak < ak + rk < ak + .
k=1 k=1 k=N k=1 k=N k=1 1r
P
This shows the partial sums of a n are bounded. Therefore, it must converge.
1/k
If > 1, there is an increasing sequence of integers k n ! 1 such that a k n > 1
P n
for all n 2 N. This shows a kn > 1 for all n 2 N. By Theorem 4.3, a n diverges.
P
E XAMPLE 4.9. For any x 2 R, the series |x n |/n! converges. To see this, note
that according to Exercise 3.3.7,
1/n
|x n | |x|
= ! 0 < 1.
n! (n!)1/n
Applying the Root Test shows the series converges.
P P
E XAMPLE 4.10. Consider the p-series 1/n and 1/n 2 . The first diverges
and the second converges. Since n 1/n ! 1 and n 2/n ! 1, it can be seen that when
= 1, the Root Test in inconclusive.
P
T HEOREM 4.11 (Ratio Test). Suppose a n is a positive series. Let
a n+1 a n+1
r = lim inf
lim sup = R.
an an
P P
If R < 1, then a n converges. If r > 1, then a n diverges.
P ROOF. First, suppose R < 1 and 2 (R, 1). There exists N 2 N such that
a n+1 /a n < whenever n N . This implies a n+1 < a n whenever n N . From
this its easy to prove by induction that a N +m < m a N whenever m 2 N. It follows
In calculus books, the ratio test usually takes the following simpler form.
P
C OROLLARY 4.12 (Ratio Test). Suppose a n is a positive series. Let
a n+1
r = lim
.
n!1 a n
P P
If r < 1, then a n converges. If r > 1, then a n diverges.
From a practical viewpoint, the ratio test is often easier to apply than the root
test. But, the root test is actually the stronger of the two in the sense that there
are series for which the ratio test fails, but the root test succeeds. (See Exercise
4.10, for example.) This happens because
a n+1 a n+1
(4.9) lim inf lim inf a n1/n lim sup a n1/n lim sup .
an an
To see this, note the middle inequality is always true. To prove the right-hand
inequality, choose r > lim sup a n+1 /a n . It suffices to show lim sup a n1/n r . As in
the proof of the ratio test, a n+k < r k a n . This implies
an
a n+k < r n+k ,
rn
which leads to
a 1/(n+k)
1/(n+k) n
a n+k <r n
.
r
Finally,
a 1/(n+k)
1/(n+k) n
lim sup a n1/n = lim sup a n+k lim sup r = r.
k!1 k!1 rn
The left-hand inequality is proved similarly.
Most times the simple tests of the preceding section suffice. However, more
difficult series require more delicate tests. There dozens of other, more spe-
cialized, convergence tests. Several of them are consequences of the following
theorem.
P
T HEOREM 4.13 (Kummers Test). Suppose a n is a positive series, p n is a
sequence of positive numbers and
an an
(4.10) = lim inf p n p n+1 lim sup p n p n+1 =
a n+1 a n+1
P P P
If > 0, then a n converges. If 1/p n diverges and < 0, then a n diverges.
P
P ROOF. Let s n = nk=1 a k , suppose > 0 and choose r 2 (0, ). There must be
an N > 1 such that
an
pn p n+1 > r, 8n N .
a n+1
Rearranging this gives
Kummers test is powerful. In fact, it can be shown that, given any positive
series, a judicious choice of the sequence p n can always be made to determine
whether it converges. (See Exercise 4.17, [20] and [19].) But, as stated, Kummers
test is not very useful because choosing p n for a given series is often difficult.
Experience has led to some standard choices that work with large classes of series.
For example, Exercise 4.9 asks you to prove the choice p n = 1 for all n reduces
Kummers test to the standard ratio test. Other useful choices are shown in the
following theorems.
P
T HEOREM 4.14 (Raabes Test). Let a n be a positive series such that a n > 0 for
all n. Define
an an
= lim sup n 1 lim inf n 1 =
n!1 a n+1 n!1 a n+1
P P
If > 1, then a n converges. If < 1, then a n diverges.
2See 5.2.
1.0
0.8
0.6
0.4
0.2
0 5 10 15 20 25 30 35
F IGURE 4.3. This plot shows the first 35 partial sums of the alternating
harmonic series. It can be shown it converges to ln 2 0.6931, which is
the level of the dashed line. Notice how the odd partial sums decrease
to ln 2 and the even partial sums increase to ln 2.
P
(a) s n = nk=1 a k is a bounded sequence.
(b) b n b n+1 , 8n 2 N
(c) b n ! 0
P
Then a n b n converges.
n
X n
X n
X k
X
a k b k = b n+1 ak (b k+1 b k ) a`
k=1 k=1 k=1 `=1
Pn
P ROOF. Let s 0 = 0 and s n = k=1
a k when n 2 N. Then
n
X n
X
ak bk = (s k s k1 )b k
k=1 k=1
Xn n
X
= sk bk s k1 b k
k=1 k=1
!
n
X n
X
= sk bk s k b k+1 s n b n+1
k=1 k=1
n
X n
X k
X
= b n+1 ak (b k+1 b k ) a`
k=1 k=1 `=1
= (b n+1 + b m )M + M (b m b n+1 )
= 2M b m
"
< 2M
2M
<"
Pn
This shows `=1
a ` b ` satisfies Theorem 4.4, and therefore converges.
Theres one special case of this theorem thats most often seen in calculus
texts.
P ROOF. Let a n = (1)n+1 and b n = c n in Theorem 4.18 to see the series con-
P
verges to some number s. For n 2 N, let s n = nk=0 (1)k+1 c k and s 0 = 0. Since
E XAMPLE 4.13. Corollary 4.20 provides another way to prove the alternating
harmonic series in Example 4.12 converges. Figures 4.3 and 4.4 show how the
partial sums bounce up and down across the sum of the series.
4. Rearrangements of Series
This is an advanced section that can be omitted.
We want to use our standard intuition about adding lists of numbers when
working with series. But, this intuition has been formed by working with finite
sums and does not always work with series.
P P
E XAMPLE 4.14. Suppose (1)n+1 /n = so that (1)n+1 2/n = 2. Its easy
to show > 1/2. Consider the following calculation.
X 2
2 = (1)n+1
n
2 1 2 1
= 21+ + +
3 2 5 3
P ROOF. (Theorem 4.23) Let b n and c n be as in Lemma 4.24 and define the
subsequence a n+ of b n by removing those terms for which b n = 0 and a n 6= 0.
Define the subsequence a n of c n by removing those terms for which c n = 0. The
P P1
series 1 +
n=1 a n and n=1 a n are still divergent because only terms equal to zero
have been removed from b n and c n .
Now, let c 2 R and m 0 = n 0 = 0. According to Lemma 4.24, we can define the
natural numbers
Xn m1
X n
X
m 1 = min{n : a k+ > c} and n 1 = min{n : a k+ + a ` < c}.
k=1 k=1 `=1
If m p and n p have been chosen for some p 2 N, then define
( ! )
Xp mXk+1 nXk+1 Xn
+ +
m p+1 = min n : a` a` + a` > c
k=0 `=m k +1 `=n k +1 `=m p +1
and
( !
p
X mX
k+1 nX
k+1
n p+1 = min n : a `+ a `
k=0 `=m k +1 `=n k +1
nX
)
p+1 n
X
+ a `+ a ` <c .
`=m p +1 `=n p +1
and !
p
X mX
k+1 nk
X
0<c a `+ a ` a np
k=0 `=m k +1 `=n k +1
+
Since both am
! 0 and
p
a n p ! 0, the result follows from the Squeeze Theorem.
The argument when c is infinite is left as Exercise 4.31.
A moral to take from all this is that absolutely convergent series are robust
and conditionally convergent series are fragile. Absolutely convergent series can
be sliced and diced and mixed with careless abandon without getting surprising
results. If conditionally convergent series are not handled with care, the results
can be quite unexpected.
5. Exercises
1
X
4.3. The series (a n a n+1 ) converges iff the sequence a n converges.
n=1
P
4.4. Prove or give a counter example: If |a n | converges, then na n ! 0.
(4.14) a1 + 0 + a2 + 0 + 0 + a3 + 0 + 0 + 0 + a4 + .
P
4.6. If 1 n=1 a n converges and b n is a bounded monotonic sequence, then
P1
n=1 a n b n converges.
P1 n
4.7. Let x n be a sequence with range {0, 1, 2, 3, 4, 5, 6, 7, 8, 9}. Prove that n=1 x n 10
converges and its sum is in the interval [0, 1].
4.9. Prove the ratio test by setting p n = 1 for all n in Kummers test.
4.12. Does
1 12 123 1234
+ + + +
3 35 357 3579
converge?
converge?
X1 (x + 3)n
4.21. For what values of x does n
converge absolutely, converge
n=1 n4
conditionally or diverge?
1
X n +6
4.22. For what values of x does converge absolutely, converge
n=1 n 2 (x 1)n
conditionally or diverge?
P1 n
4.23. For what positive values of does n=1 n converge?
X n
4.24. Prove that cos sin converges.
3 n
P1
4.25. For a series k=1
a n with partial sums s n , define
1 Xn
n = sn .
n k=1
P
Prove that if 1 a = s, then n ! s. Find an example where n converges, but
k=1 n
P1
k=1
a n does not. (If n converges, the sequence is said to be Cesro summable.)
P
4.26. If a n is a sequence with a subsequence b n , then 1 n=1 b n is a subseries
P P1 P
of 1n=1 na . Prove that if every subseries of a
n=1 n converges, then 1 n=1 a n
converges absolutely.
P P1 2
4.27. If 1 n=1 a n is a convergent positive series, then so is n=1 a n . Give an
example to show the converse is not true.
P1 P1 2
4.28. Prove or give a counter example: If n=1 a n converges, then n=1 a n
converges.
P1 n
4.29. For what positive values of does n=1 n converge?
4.30. If a n 0 for all n 2 N and there is a p > 1 such that limn!1 n p a n exists and
P
is finite, then 1 n=1 a n converges. Is this true for p = 1?