Вы находитесь на странице: 1из 7

ANALYSIS I

9 The Cauchy Criterion

9.1 Cauchy’s insight

Our difficulty in proving “a n ” is this: What is ? Cauchy saw that it was enough to show that if the terms of the sequence got sufficiently close to each other. then completeness will guarantee convergence.

Remark. In fact Cauchy’s insight would let us construct R out of Q if we had time.

9.2 Definition

Let ( a n ) be a sequence [R or C ]. We say that ( a n ) is a Cauchy sequence if, for all ε > 0 there exists N N such that

m, n N =⇒ |a m a n | < ε.

[Is that all? Yes, it is!]

9.3 Cauchy =Bounded

Theorem. Every Cauchy sequence is bounded [R or C ].

Proof. 1 > 0 so there exists N such that m, n N =⇒ | a m a n | < 1. So for m N , |a m | 1 + |a N | by the ∆ law. So for all m

|a m | 1 + |a 1 | + |a 2 | + ··· + |a N |.

 
 

9.4

Convergent =Cauchy [ R or C ]

 

Theorem. Every convergent sequence is Cauchy.

 

Proof. Let a n l and let ε > 0. Then there exists N such that

 
 

k N =⇒ |a k l | < ε/2

For m, n N we have

 
 

|a m

l |

< ε/ 2

|a n l |

< ε/ 2

So

 

|a m a n |

|a m l |

+

|a n l |

by the ∆ law

 

<

ε/2

+

ε/ 2

= ε

 
 

1

9.5

Cauchy =Convergent [ R ]

Theorem. Every real Cauchy sequence is convergent.

Proof. Let the sequence be (a n ). By the above, (a n ) is bounded. By Bolzano-Weierstrass (a n ) has a convergent subsequence (a n k ) l , say. So let ε > 0. Then

N 1 such that N 2 such that

ε/2

m, n N 2 =⇒ | a m a n | < ε/2

r N 1 =

|a n r l |

<

Put s := min { r |n r N 2 } and put N = n s . Then

m, n N

=

| a m a n |

|a m a n s |

+

|a n s l |

<

ε/2

+

ε/ 2

= ε

s − l | < ε/ 2 + ε/ 2 = ε 9.6 Cauchy = ⇒

9.6 Cauchy =Convergent [ C ]

Theorem. Every complex Cauchy sequence is convergent.

Proof. Put z n = x + i y .

x n x, y n y and so z n x + iy .

Then x n is Cauchy:

|x x x m | |z n z m | (as | w | |w |).

9.7 Example

Let

1

1

3 + ··· + (1) n+1

2 +

Then with m n , and m n odd we have

a n = 1

n

1

1

1

1

+

1

1

|a m a n |

=

=

n + 1 n + 2

1

1

n + 1

+

n + 3 n + 4 +

1

1

m 1 m

1

1

n + 2 + n + 3

m

m 2 + m 1

1

1

n+1 n

If m n is even, we write

|a m a n | = |a m a m 1 + a m | |a m a m 1 | + |a m |

1

Let ε > 0. Chose N > ε and convergence follows. Question: What is lim n a n ? Later we’ll see it is log 2.

2

2

N + 1 + m n

1

1

2

So

follows. Question: What is lim n → ∞ a n ? Later we’ll see it is

10 Series

10.1 Definition

Note. So far we have always used sequences defined by functions a : N R . It is convenient from now on to start off at a 0 , that is to work with functions a : N ∪ {0} → R .

So let (a n ) n=0 be a sequence of real (complex) numbers. Define s m := n=0 m a n , a well defined real (complex) number.

Consider the sequence (s n ) n=0 . We call this the series defined by the sequence ( a n ), and we

denote it by

n=0 a n .

Note. This is a very odd name indeed! Don’t let it mislead you: n=0 a n is just the sequence a 0 , a 0 + a 1 , a 0 + a 1 + a 2 ,

10.2 Examples

(i) The Geometric Series Let a n := x n Then x n is

(1, 1 + x, 1 + x + x 2 ,

, 1 + x + x 2 + ··· + x n ,

)

1

(ii) The Harmonic Series Let a n := n+1 . Then

1, 1 + 2 , 1 +

1

1

2 +

1

n+1 is

1

3 ,

.

.

.

(iii)

The Exponential Series Let a n := x n /n !. Then x n /n ! is

1 , 1 + x, 1 + x, 1 + x + x 2! 2 ,

(iv)

The Cosine Series Let

Then a n is

a n =

x 2 m

2 m!

0

(1) m if n = 2m otherwise

1, 1, 1 x 2 , 1 x 2 , 1 x 2 + x 4 ,

2!

2!

2!

4!

10.3 Convergence and Divergence

With the setup above if (s n ) converges, then we say “the series n=0 is convergent ”. Otherwise we say that the series is divergent .

3

10.4 More examples

(i) Let a n = x n .

(a)

If x = 1 then x n is divergent.

Proof. s 0 = 1, s 1 = 2,

, s n = n + 1 by induction.

(b) If x = 1, then s n = 1 x n+1

1 x

.

Proof. Pure algebra.

(c)

(d)

If |x| < 1 then x n is convergent.

If |x| > 1 then x n is divergent.

Proof.

then x n+1 11 (1 x)s by AOL. But if |x| > 1, x n is not convergent.

If |x| < 0 then x n 0 as n → ∞.

1

1

So by AOL s n 1/ (1 x). If s n s

8 4

7 + 8 +

+

1

1

.

.

.

.

.

.

+

2 n1

2

n

(ii) Let a n = n . Then n is divergent.

Proof.

s 2 n 1 = 1 + 1 2 + ··· +

=

1

1

n

2

+

+

1

2

1 2

+

+

1

2 n =

1

3 +

2

4

1

4

+

+

1

5 + 1 6 +

so (s n ) has a subsequence which is not convergent.

(iii) Let a n =

Proof.

s n =

=

=

1

( n+1) 2 . Then

2

1

1

1

1

n+1

2

+

+

+

1

2 is convergent.

n

1

1

1

4

1

1 · 2

1 2

+

+

+

1

9

1

+

3 +

+

2 ·

1

2

1

3

.

.

.

.

.

.

.

.

.

+

+

+

1

n

2

1

n( n+1)

1

n

n+1

1

+

1

(

n+1) 2

So ( s n ) is monotone increasing [ s n+1 = s n + ( n+1) 2 s n ] and bounded by 2. So (s n ) is

convergent by Bolzano-Weierstrass ( ?? ).

1

+ ( n + 1 ) 2 s n ] and bounded by 2. So (

4

10.5 Notation and its abuse

More notation: if the series n=0 a n is convergent then we often denote the limit by and call it the sum.

n=0 a n ,

Note. We must take great care, but this double use is traditional. I will try to distinguish the two uses, and say “The series a n ” and “The sum a n ”. I suggest that you do the same for a bit.

Note. Just attaching the label “sum” to the symbol n=0 a n does not turn it into a proper mathematical sum. Look at our axioms for R . They only speak of adding pairs of real numbers, which we can extend using axiom A2 to finite sets of real numbers. But given an infinite set of real numbers we can’t simply “add” them and get a “sum”. Instead we have to talk about sequences

10.6 Tails

Let (a n ) be a sequence and (s n ) be the corresponding series.

Sometimes we want to look at (a k , a k +1 , a k +2 ,

S n = a k + a k +1 + ··· + a n and note that what we said about Tails of sequences, and adding constants, ensures that ( s n ) is convergent if and only if S n is convergent.

). We write this series n= k a n . We put

10.7 Cauchy’s criterion

We rewrite Cauchy Criterion for series.

Theorem. The series n=0 a n is convergent if and only if for all ε > 0 there exists N N such that

l k > N

=

l

n= k

a n

A genuine sum

< ε

Note. Clearly in practice when we estimate the sum we’ll use the law when we can.

10.8 Absolute Convergence

Let a n be a sequence. Then we say that a n is absolutely convergent if the series |a n | is convergent.

10.9 Absolute Convergence =Convergence

Theorem. If a n is absolutely convergent then it is convergent.

Proof. Let ε > 0 Then there exists N N such that

l k N =

l

k

|a n | < ε

5

So

l k N =

l

k

l

k

|a n | =

a n

By the ∆ law

l

k

|a n | < ε

| = a n By the ∆ law l k | a n | < ε

Note. Why is absolute convergence a good thing to have? Because it makes use of Cauchy criterion easy!

10.10 Examples

(i) x n absolutely convergent if |x| < 1

(ii) ( 1) n is absolutely convergent

n

2

(iii)

sin n is absolutely convergent

n

3

(iv) ( 1) n

n+1 is convergent, but not absolutely convergent.

10.11 Re-arrangements

Let p : N −→ N one-to-one and onto. We can then put b n = a p( n) and consider b n , which we call a rearrangement of the series a n . Funny this can happen! Later on we will be able to prove that

Example.

1 1 2 + 1 + 1 3 +

1

3

1

5

1

+ 2 +

4

1

.

1

7 + 1 9 + 11

.

.

1

1

4

+

.

.

.

log(2/3)

log 2

Theorem. If (a n ) is absolutely convergent and (b n ) is a rearrangement of (a n ) then b n is absolutely convergent too.

10.12 Multiplication of series

Theorem. Suppose a n , b n are absolutely convergent. Suppose that

Then

(i)

(ii)

c n

c n = a n b n

is absolutely convergent

Proof.

c n :=

n

r =0

6

a r b nr

(i) By pure algebra

N

0

|a n |

N

0

|b n |

2

0

N |c r |

2 N

|a n |

2 N

|b n |

0 0

So by Sandwich Rule, |c n | is convergent to |a n | |b n |.

(ii) Hence c n is convergent.

(iii) Given ε > 0 there exist N such that

Then for k N

2

k

n=0

c n

k

n=0

a

n

k

b

n

n=0

l k N =

=

r + s 2 k, r k or

l

k

s k

|c n | < ε

a r b s

ditto

|a r b s |

2

k

k

|c n | < ε

b s ditto | a r b s | 2 k k | c n |

Note. Draw a diagram in the (r, s) plane marking out which a r b s are included in the various sums.

10.13 Two Applications

(i)

(n + 1)x n =

1

(1 x) 2

Proof. a n := x n , b n := x n , therefore

c n = x r x s = (n + 1)x n

(ii)

Proof.

r + s = n

x n y n = (x + y ) n

n

!

n

!

n

!

a n := x n , b n := y n therefore

n!

n!

c n =

r + s = n

x r y s = (x + y ) n

r ! s!

n

!

7