13.4.sterling’s series derivation from euler-maclaurin integration formula euler-maclaurin...

25
13.4. Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : 2 1 2 1 2 0 1 0 1 0 0 2 2 ! n q n p p p q m p B dx f x f m n f n f R p 2 1 f x z x B 2 B 4 B 6 B 8 1/ 6 1/3 0 1/4 2 1/3 0 Let 0 0 1 dx f x z x 1 z 1 2 0 1 1 0 1 2 2 m f m z z 1 1 1 1 ! 1 m m m k z m z k 2 1 2 1 2 ! p p p f x z x 1 3 2 f x z x 3 5 234 f x z x 1 2 2 2 1 1 1 1 1 2 p p p B z z z z

Upload: hortense-warren

Post on 18-Jan-2016

263 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

13.4. Sterling’s SeriesDerivation from Euler-Maclaurin Integration Formula

Euler-Maclaurin integration formula :

2 1 2 12

0 10

10 0

2 2 !

n qnp pp

qm p

Bd x f x f m f f n f n f R

p

2

1f x

z x

B2 B4 B6 B8

1/6 1/30 1/42 1/30

Let 00

1d x f x

z x

1

z

1

20

1 10 1

2 2m

f m f f zz

1

11

1!

1

mmm

k

z mz k

2 12 1

2 !pp

pf x

z x

13

2f x

z x

3

5

2 3 4f x

z x

1 2

2 2 11

1 11

2p

pp

Bz

z z z

Page 2: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

1 2

2 2 11

1 11

2p

pp

Bz

z z z

1 2

2 2 11

1 11

2p

pp

Bz

z z z

111 1z C d z z 2

1 21

1ln

2 2p

pp

BC z

z p z

2ln 1 1z C d z z

2

2 2 11

1ln ln

2 2 2 1p

pp

BC z z z z

p p z

1lim 1 lnz

z C z

1

1 1

1k

zz k k

1

1

k k

1 0C

2

1lim ln 1 ln

2zz C z z z

Page 3: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

2

11 2 1

2 2 zz z z

1ln 2 ln 2

2z

2

1ln 2

2C

2

1lim ln 1 ln

2zz C z z z

2

1 1 1lim ln ln

2 2 2zz C z z z

1 1ln ln

2 2z z

z

2

1lim ln ln

2zz C z z z

2

1lim ln 2 1 2 2 ln 2

2zz C z z z

2

1 1ln 2 2 2ln 2 2 ln

2 2C z z z

2

11

12lim ln ln 2 2ln 22 1 2z

z zC z

z

2

2 11

1 1ln 1 ln 2 ln

2 2 2 2 1p

pp

Bz z z z

p p z

Stirling’s series

Page 4: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

2

2 11

1 1ln 1 ln 2 ln

2 2 2 2 1p

pp

Bz z z z

p p z

z >> 1 : 1 1ln 1 ln 2 ln

2 2z z z z

1/2ln 2 z zz e 1/21 2 z zz z e

Stirling approx

A = Arfken’s two-term approx. using1/12 1

112

zez

Mathematica

Page 5: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

13.5. Riemann Zeta Function

Riemann Zeta Function : 1

1z

n

zn

2 4 6 8 10

" " 2 3 4 5 6 7 8 9 10

" ( )" 1.202 1.037 1.008 1.0026 90 945 9450 93555

z

z

Integral representation : 1

0

1

1

z

t

tz d t

z e

Proof :

1

0

1

1

tz

t

eRHS d t t

z e

1

1 0

1 z mt

m

d t t ez

1

1 0

1 1 z sz

m

d s s ez m

1

1z

m m

s m t

LHS

Mathematica

Page 6: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

Definition : Contour Integral

1

11

1

z

tC

tI d t

z e

11

1

z

A x

xI d x

z e

z

1

0

1

1

z

t

tz d t

z e

12

1

1

zi

B x

x eI d x

z e

2 1z iz e 2 z iz e

1

2

0

1

1i

zi

iD e

eI i e d

z e

2

21

0

1 i zzi e dz

0 for Re z >1

diverges for Re z <1

1

1

2

1

11

z

tz i C

tz d t

ee z

agrees with integral

representation for Re z > 1

C1

Page 7: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

Similar to ,

Definition valid for all z (except for z integers).

Analytic Continuation

1

1

2

1

11

z

tz i C

tz d t

ez e

Poles at 2 0, 1, 2,t n i n

1

1 1

1 1C

z

t C

z

t

t td t d te e

1

11

2 21

zz

C tn

td t i n ie

Re z > 1

1

1 1 1

2 Res ; 21 1 1C

z z

C

z

t t tn

t t td t d t i n ie e e

1/ 2 1

1

2 1z i zi z

n

e e n

C C1 encloses no pole.C C1 encloses all poles.

means n 0

/ 2 3 / 22 1z i z i ze e z

3 / 2 / 2

2

21

1

z i z i z

z i

e ez z

z e

11

1

2z zz

n

i n n

Mathematica

Page 8: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

Riemann’s Functional Equation

3 / 2 / 2

2

21

1

z i z i z

z i

e ez z

z e

3 / 2 / 2 / 2 / 2

2 1

i z i z i z i z

z i z i z i

e e e e

e e e

sin2

sin

z

z

1

/ 2 1 / 2

z z

z z

1sin

z zz

12 1 sin 12

z zz z z z

1sin 1

2z z z

Riemann’s functional equation

Page 9: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

Zeta-Function Reflection Formula

3 / 2 / 2

2

21

1

z i z i z

z i

e ez z

z e

3 / 2 / 2

2

1

1 / 2 1 / 2

i z i z

z i

z ze e

e z z

2 1

1/ 2 1 / 2

zz

z zz z

2

11 2 1

2 2 zz z z

11 1

2 2 2 2 z

z zz

1/2 1

2 1/ 2

z z

z zz

zeta-function reflection formula

1 /2/2 1/ 2 1

2zz z

z z z

Page 10: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

12 1 sin 12

z zz z z z

Riemann’s functional equation :

for trivial zeros 0z 2z n 1, 2, 3,n

1

1z

n

zn

converges for Re z > 1

12 1 sin 12

z zz z z z

(z) is regular for Re z < 0.

(0) diverges (1) diverges while (0) is indeterminate.

Since the integrand in is always positive,

(except for the trivial zeros)

or

i.e., non-trivial zeros of (z) must lie in the critical strip

1

0

1

1

z

t

tz d t

z e

0 Re 1z z

1 0 Re 1z z

0 Re 0z z

0 Re 1z

Page 11: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

Critical Strip

1

1

n

zn

zn

Consider the Dirichlet series :

Leibniz criterion series converges if , i.e., 1

lim 0zn n Re 0z

1 1

1 1

2 1 2z z

n n

zn n

1 1

1 1 12

2z z zn nn n

11 2 z z

for 11 2 z

zz

Re 0z

11

1Res ;1 lim

1 2 zz

z zz

1 ln 212 zz e

11

ln 2

1

122 ln 2

zzd

d z

1 ln 2 1

1

ln 1n

n

xx

n

1

1

1z

n

zn

Page 12: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

(0)

2 1

1/ 2 1 / 2

zz

z zz z

0

10 lim

/ 2z

z

z

10

2

0

Res ;1

1 1lim

Res ; 0

02

z

s

z

sz

Simple poles :

Res ;11

2 Res ; 0

s

s

Res ;1 1z

1

limz n

nk

n kz

z z k

Res ; 0 1z

Page 13: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

Euler Prime Product Formula

1 1 1 1 1 11 1

2 3 5 7 9 11s s s s s ss

1

1z

n

zn

1 1 1 1 1 1 1 1 1 11

2 3 4 5 6 7 8 9 10 11s s s s s s s s s ss

( no terms ) 1

2s

n

1 1 1 1 11 1 1

2 3 5 7 11s s s s ss

( no terms ) 1

3s

n

primes

11 1

sp

sp

primes

1

1 sp

sp

Euler prime product formula

Page 14: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

Riemann Hypothesis

Riemann found a formula that gives the number of primes less than a

given number in terms of the non-trivial zeros of (z).

Riemann hypothesis :

All nontrivial zeros of (z) are on the critical line Re z ½.

Millennium Prize problems proposed by the Clay Mathematics Institute.

1. P versus NP

2. The Hodge conjecture

3. The Poincaré conjecture (proved by G.Perelman in 2003)

4. The Riemann hypothesis

5. Yang–Mills existence and mass gap

6. Navier–Stokes existence and smoothness

7. The Birch and Swinnerton-Dyer conjecture

Page 15: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

13.6. Other Related Functions

1. Incomplete Gamma Functions

2. Incomplete Beta Functions

3. Exponential Integral

4. Error Function

, , ,a x a x

Page 16: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

Incomplete Gamma Functions

1

0

,x

t aa x d t e t

1, t a

x

a x d t e t

, ,a x a x a

Integral representation:

Re 0a

0,t

x

ex d t

t

Ei x 0x Exponential integral

Page 17: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

Series Representation for (n, x)

1

1 0

1 !1 !

!

xnx n k t

k

ne x n d t e

n k

1

0

,x

t nn x d t e t 1 2

00

1x

xt n t ne t n d t e t

1 2 3

0

1 2x

x n x n t ne x n e x n d t e t

1

1

11 ! 1

!

nx n k x

k

n e x en k

1

11 ! 1

!

nx n k

k

n e xn k

1

0

1, 1 ! 1

!

nx s

s

n x n e xs

s n k

1, 2, 3,n

Page 18: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

Series Representation for (n, x)

1

1

1 !1 !

!

nx n k t

k x

ne x n d t e

n k

1, t n

x

n x d t e t

1 21t n t n

xx

e t n d t e t

1 2 31 2x n x n t n

x

e x n e x n d t e t

1

1

11 !

!

nx n k x

k

n e x en k

1

11 !

!

nx n k

k

n e xn k

1

0

1, 1 !

!

nx s

s

n x n e xs

s n k

1, 2, 3,n

Page 19: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

Series Representation for (a, x) & (a, x)

For non-integral a :

0

,!

n

a n

n

a x x xn a n

0x

See Ex 1.3.3 & Ex.13.6.4

1

0

1, ~ a x

nn

aa x x e

a n x

x

1

0

1a xnn

n

x e a nx

Pochhammer symbol

1 1n

a a a a n

01a

1

0

10, ~ x

nnn

x x e nx

0

!xn

nn

e n

x x

1a a

Relation to hypergeometric functions: see § 18.6 .

Page 20: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

Incomplete Beta Functions

1

11

0

, 1qpB p q d t t t

11

0

, 1x

qpxB p q d t t t

0 ,1 & 0x p

, 0p q

0

1

!p nn

n

qx x

n p n

Ex.13.6.5

Relation to hypergeometric functions: see § 18.5.

Page 21: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

Exponential Integral Ei(x)

0,t

x

ex d t

t

Ei x 1E x

t

x

eEi x P d t

t

x te

P d tt

0x

0x

1 0E

P = Cauchy principal value

1

t x

n n

eE x d t

t

E1 , Ei analytic continued.

Branch-cut : (x)–axis.

1 0E x i Ei x i

1 1

10 0

2Ei x E x i E x i

Mathematica

Page 22: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

Series Expansion

1 0lim ,a

E x a a x

0

1

lim!

naa n

an

xa x x

a n a n

0

1

lim!

nan

an

a a xx

a n a n

0

,!

n

a n

n

a x x xn a n

0

0 0

1 1lim lim

a a

a a

a a x a x x

a a a

01

z

zz

d z d x

d z d z

1 1 ln x

ln lnz

z z x zd xx e x x

d z

lnd

z zd z

ln x

1

1

ln!

n

n

n

E x x xn a n

For x << 1 :

For x >> 1 : 10

!0,

xn

nn

e nE x x

x x

Page 23: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

Sine & Cosine Integrals

0

0

0

sin sin

cos 1 cosln

lnln

x

x

x

x

x

t tSi x d t si x d t

t t

t tCi x x d t ci x d t

t t

d tli x P Ei x

t

0

sin

2

tSi x si x d t

t

Ci(z) & li(z) are multi-valued.

Branch-cut : (x)–axis.

1li

0

1 cosx tCin x d t

t

is an entire function

0

cosx td t

tnot defined

Mathematica

Page 24: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

1

cos

sin

x

x

t

x

t

x

tCi x d t

t

tsi x d t

t

eEi x d t

t

eE x d t

t

t s x

1

2Ei ix Ei i x

i

1

1

s xeE x d s

s

t

x

eEi x d t

t

1

sin sxsi x d s

s

1

1

2

i s x i s xe esi x d s

i s

1

cos sxCi x ci x d s

s

1 1

1

2E ix E ix

i

1

s xed s

s

1

1

2

i s x i s xe eci x d s

s

1

2Ei ix Ei i x 1 1

1

2E ix E ix

Ei ix ci x i si x 1E ix ci x i si x

Series expansions : Ex.13.6.13. Asymptotic expansions : § 12.6.

Page 25: 13.4.Sterling’s Series Derivation from Euler-Maclaurin Integration Formula Euler-Maclaurin integration formula : B2B2 B4B4 B6B6 B8B8 1/6  1/30 1/42

Error Function

2

0

2 zterf z d t e

221 t

z

erfc z erf z d t e

1erf

Power expansion :

2

0 0

2

!

n xn

n

erf x d t tn

2 1

0

2

! 2 1

n

n

n

xn n

Asymptotic expansion (see Ex.12.6.3) :

221 t

x

erf x d t e

2 2

2

2 11

2 2

x t

z

e ed t

x t

2

2

2

tt de

e d tt

2

2 10

2 1 !!1

2

nx

n nn

ne

x

21 1,

2z

1

0

,x

t aa x d t e t

21 1,

2z

Mathematica