c2999 PDF Supa

Embed Size (px)

Citation preview

  • 7/28/2019 c2999 PDF Supa

    1/26

    Supplement A

    Special Functions and Their Properties

    Throughout Supplement A it is assumed that#

    is a positive integer, unless otherwise specified.

    A.1. Some Symbols and Coefficients

    A.1.1. Factorials

    Definitions and some properties:

    0! = 1! = 1,#

    ! = 1 2 3 33 3

    (#

    1)#

    ,#

    = 2, 3, 33 3

    ,

    (2 # )!! = 2 4 6 3 3 3 (2 # 2)(2 # ) = 2 # !,

    (2#

    + 1)!! = 1 3 5 33 3

    (2#

    1)(2#

    + 1) =2 +1

    7 8 9 @

    #+

    3

    2A

    ,

    #!! =

    B

    (2C

    )!! if #

    = 2C

    ,

    (2 C + 1)!! if # = 2 C + 1,0!! = 1.

    A.1.2. Binomial Coefficients

    Definition:

    4 D

    E =# !

    C!(

    #

    C)!

    , whereC

    = 1, 33 3

    ,#

    ,

    4 D

    F = (1)

    D

    (G

    )

    D

    C !=

    G(

    G

    1) 33 3

    (G

    C

    + 1)

    C !, where

    C

    = 1, 2, 33 3

    General case:

    4 H

    F =9 (

    G+ 1)

    9 (I

    + 1) 9 (G

    I

    + 1), where 9 (

    P) is the gamma function.

    Properties:

    4

    0F = 1,

    4 D

    E = 0 forC

    = 1, 2, 33 3

    orC

    >Q

    ,

    4 H

    +1F =

    G

    I

    + 14 H

    F

    1=

    G

    I

    I

    + 14 H

    F ,

    4 H

    F +

    4 H

    +1F =

    4 H

    +1F

    +1,

    4

    E

    1R2

    =(1)

    E

    22E S

    E

    2E = (1)

    E (2 Q 1)!!

    (2Q

    )!!,

    S

    E

    1 R 2=

    (1)E

    1

    Q

    22E

    1S

    E

    1

    2E

    2 =(1)

    E

    1

    Q

    (2Q

    3)!!

    (2Q

    2)!!,

    S

    2E

    +1E

    +1R2

    = (1)E

    24E

    1S

    E

    2E ,

    S

    E

    2E

    +1R2

    = 22E

    S

    2E

    4E

    +1,

    S

    1R

    2E = 2

    2E

    +1

    8

    S

    E

    2E

    ,S

    E

    R

    2E = 2

    2E

    8

    S

    (E

    1) R 2E .

    2002 by Chapman & Hall/CRC

  • 7/28/2019 c2999 PDF Supa

    2/26

    A.1.3. Pochhammer Symbol

    Definition and some properties (C

    = 1, 2, 33 3

    ):

    (G

    ) E =G

    (G

    + 1) 33 3

    (G

    +Q

    1) =9 (

    G

    +Q

    )

    9 ( G )= (1)

    E9 (1

    G)

    9 (1 G Q ),

    (G

    )0 = 1, ( G ) E +

    D

    = (G

    ) E (G

    +Q

    )

    D

    , (Q

    )

    D

    =(

    Q+

    C 1)!

    (Q

    1)!,

    (G

    ) E =9 (

    G

    Q)

    9

    (G

    )

    =(1)

    E

    (1 G

    )E

    , whereG

    1, 33 3

    ,Q

    ;

    (1) E = Q !, (1 T 2) E = 22E (2

    Q)!

    Q!

    , (3 T 2) E = 22E (2

    Q

    + 1)!

    Q!

    ,

    (G

    +U C

    ) E

    D

    =( G ) V

    D

    +E

    D

    (G

    )V

    D

    , (G

    +Q

    ) E =( G )2 E

    (G

    ) E, (

    G+

    Q)

    D

    =( G )

    D

    ( G + C ) E

    (G

    ) E.

    A.1.4. Bernoulli Numbers

    Definition:

    P

    W X

    1

    =Y

    `

    E

    =0 a

    E

    P

    E

    Q

    !

    .

    The numbers:

    a

    0 = 1,

    a

    1 = 1

    2,

    a

    2 =1

    6,

    a

    4 = 1

    30,

    a

    6 =1

    42,

    a

    8 = 1

    30,

    a

    10 =5

    66, 3

    3 3,

    a

    2 V +1 = 0 for U = 1, 2, 3 3 3

    A.2. Error Functions and Exponential Integral

    A.2.1. Error Function and Complementary Error Function

    Definitions:

    erfP

    =2

    78 b

    X

    0

    exp(c

    2)d c

    , erfcP

    = 1 erfP

    =2

    78 b

    Y

    X

    exp(c

    2)d c

    .

    Expansion of erfP

    into series in powers ofP

    asP e

    0:

    erfP

    =2

    78

    Y

    `

    D

    =0

    (1)

    D

    P

    2

    D

    +1

    (C

    )!(2C

    + 1)=

    27

    8 expf

    P

    2g

    Y

    `

    D

    =0

    2

    D

    P

    2

    D

    +1

    2C

    + 1)!!.

    Asymptotic expansion of erfcP

    asP e h

    :

    erfcP

    =1

    7 8 expf

    P

    2g i p

    1`

    V =0

    (1)V

    f

    1

    2

    g

    V

    P

    2 V +1+

    q f|

    P|2

    p

    1g r ,

    t= 1, 2, 3

    3 3

    A.2.2. Exponential Integral

    Definition:

    Ei(P

    ) = b

    X

    Y

    W u

    c

    d cfor

    P< 0,

    Ei(P

    ) = limv w

    +0 x

    b

    v

    Y

    W u

    c

    d c

    + b

    X

    v

    W u

    c

    d c yfor

    P> 0.

    2002 by Chapman & Hall/CRC

  • 7/28/2019 c2999 PDF Supa

    3/26

    Other integral representations:

    Ei( P ) = W X

    b

    Y

    0

    P sin c + c cos c

    P

    2 +c

    2d c for P > 0,

    Ei(P

    ) = W X

    b

    Y

    0

    Psin

    c

    ccos

    c

    P

    2 +c

    2d c

    forP

    < 0,

    Ei(P

    ) = P

    b

    Y

    1

    W X u lnc d c

    forP

    > 0.

    Expansion into series in powers of P as P e 0:

    Ei(P

    ) =

    + ln(P

    ) +Y

    `

    D

    =1

    P

    D

    C

    C

    !if

    P< 0,

    + lnP

    +Y

    `

    D

    =1

    P

    D

    C

    C

    !if

    P> 0,

    where

    = 0.5772 33 3

    is the Euler constant.

    Asymptotic expansion asP e h

    :

    Ei(P

    ) = W X

    E

    `

    D

    =1

    (1)

    D

    (C

    1)!

    P

    D

    +

    E ,

    E 0,

    9

    (k

    ) =l

    0m n

    1

    m

    .

    For (

    + 1) < Rek

    <

    , where

    = 0, 1, 2, 33 3

    ,

    ( k ) = l

    0

    =0

    (1)

    ! m n

    1

    m

    .

    A.4.1-2. Some formulas.

    Euler formula

    (k

    ) = lim

    z

    ! n

    k(

    k+ 1) 3

    3 3(

    k+

    )

    (k

    0, 1, 2, 33 3

    ).

    Simplest properties:

    ( k + 1) = k

    ( k ),

    ( + 1) = !,

    (1) =

    (2) = 1.

    Symmetry formulas:

    ( k )

    ( k ) = {k sin(

    {

    k ),

    ( k )

    (1 k ) = {sin(

    {

    k ),

    |

    12

    +k }

    |

    12

    k }

    ={

    cos({

    k)

    .

    Multiple argument formulas:

    (2k

    ) =22

    n

    1

    {

    (k

    ) |

    k

    +1

    2}

    ,

    (3k

    ) =33

    n

    1 ~ 2

    2{

    (k

    ) |

    k+

    1

    3}

    |

    k+

    2

    3}

    ,

    ( k

    ) = (2 { )(1

    ) ~ 2

    n

    1~2

    1

    =0

    |

    k

    +

    }

    .

    Fractional values of the argument:

    | 1

    2}

    =

    {

    ,

    |

    1

    2}

    = 2

    {

    ,

    |

    +1

    2}

    =

    {

    2

    (2

    1)!!,

    | 1

    2

    }= (1)

    2

    {

    (2 1)!!.

    Asymptotic expansion (Stirling formula):

    (k

    ) = 2{

    n

    k

    n

    1~2

    1 + 112

    k

    1 + 1288

    k

    2 +

    (k

    3)

    (|arg |k

    0):

    (k

    ) =l

    0

    (1 +m

    )

    n

    m

    1

    m

    ,

    (k

    ) = lnk

    +l

    0

    m

    1 (1 )1

    n

    m

    ,

    (k

    ) =

    +l

    1

    0

    1 m

    n

    1

    1 m

    m

    ,

    where

    =

    (1) = 0.5772 33 3

    is the Euler constant.

    Values for integer argument:

    (1) = ,

    ( ) = +

    1

    =1

    1 ( = 2, 3, 3 3 3 )

    A.4.2. Beta Function

    Definition:

    (

    ,

    ) =l

    1

    0m

    1(1 m

    )

    1

    m

    ,

    where Re

    > 0 and Re

    > 0.

    Relationship with the gamma function:

    (

    ,

    ) =

    (

    )

    (

    )

    ( + ) .

    A.5. Incomplete Gamma and Beta Functions

    A.5.1. Incomplete Gamma Function

    Definitions (integral representations):

    ( , ) = l 0

    m

    1

    m

    , Re > 0,

    (

    ,

    ) =l

    m

    1

    m

    =

    (

    )

    (

    ,

    ).

    2002 by Chapman & Hall/CRC

  • 7/28/2019 c2999 PDF Supa

    7/26

  • 7/28/2019 c2999 PDF Supa

    8/26

  • 7/28/2019 c2999 PDF Supa

    9/26

  • 7/28/2019 c2999 PDF Supa

    10/26

  • 7/28/2019 c2999 PDF Supa

    11/26

    2m

    . Letn

    =n

    be positive zeros of the Bessel function derivative z

    (n

    ), where

    > 1 and

    = 1, 2, 3,

    Then the set of functions

    (n

    o

    k

    ) is orthogonal on the interval 0 o

    with

    weight o :

    0

    n

    o

    n

    o

    o o = |

    }~

    0 if

    ,

    1

    2

    2

    1

    2

    n

    2

    2 ( n ) if = .

    3 m . Let n = n be positive roots of the transcendental equationn z

    ( n ) +

    ( n ) = 0, where > 1

    and = 1, 2, 3,

    Then the set of functions

    ( n ok

    ) is orthogonal on the interval 0 o

    with weight o :

    0

    n

    o

    n

    o

    o o =|

    }~

    0 if ,

    1

    2

    2

    1 +

    2 2

    n

    2

    2 ( n ) if = .

    4 m . Let n = n be positive roots of the transcendental equation

    (

    )

    (

    )

    (

    )

    (

    ) = 0 (

    > 1,

    = 1, 2, 3,

    ).

    Then the set of functions

    (

    o ) =

    (

    o )

    (

    )

    (

    )

    (

    o ),

    = 1, 2, 3,

    ,

    satisfying the conditions

    ( ) =

    ( ) = 0 is orthogonal on the interval o with

    weight o :

    ( o )

    (

    o ) o o = |

    }~

    0 if ,2

    2

    2

    2 ( ) 2

    ( )

    2

    ( )if = .

    5 m . Let n = n be positive roots of the transcendental equation

    z

    (

    )

    z

    (

    )

    z

    (

    )

    z

    (

    ) = 0 (

    > 1,

    = 1, 2, 3,

    )

    Then the set of functions

    ( o ) =

    ( o )

    z

    ( ) z

    ( )

    ( o ), = 1, 2, 3,

    ,

    satisfying the conditions

    z

    (

    ) =

    z

    (

    ) = 0 is orthogonal on the interval

    o

    withweight o :

    ( o )

    (

    o ) o o = |}

    ~

    0 if ,

    2 2

    2

    1

    2

    2

    2

    z

    ( ) {2

    z

    (

    ){

    2

    1

    2

    2

    2

    if = .

    A.6.4. Hankel Functions (Bessel Functions of the Third Kind)

    The Hankel functions of the first kind and the second kind are related to Bessel functions by (1) ( ) =

    ( ) +

    ( ), (2) ( ) =

    ( )

    ( ), 2 = 1.

    Asymptotics for x

    0:

    (1)0

    ( ) 2

    ln , (1) ( )

    ( )

    ( k

    2) (Re > 0),

    (2)0

    ( ) 2

    ln , (2) ( )

    ( )

    ( k

    2) (Re > 0).

    Asymptotics for | |x y

    :

    (1) ( ) 2

    exp

    12

    14

    ( < arg < 2 ),

    (2) (

    )

    2

    exp

    12

    1

    4

    (2 < arg

    < ).

    2002 by Chapman & Hall/CRC

  • 7/28/2019 c2999 PDF Supa

    12/26

  • 7/28/2019 c2999 PDF Supa

    13/26

  • 7/28/2019 c2999 PDF Supa

    14/26

    A.8. Airy Functions

    A.8.1. Definition and Basic Formulas

    A.8.1-1. The Airy functions of the first and the second kinds.

    The Airy function of the first kind, Ai(

    ), and the Airy function of the second kind, Bi(

    ), are

    solutions of the Airy equation

    = 0

    and are defined by the formulas

    Ai(

    ) =1

    0

    cos 13

    3 +

    ,

    Bi(

    ) =1

    0

    exp 13

    3 +

    + sin 13

    3 +

    .

    Wronskian:

    Ai(

    ), Bi(

    ) = 1

    .

    A.8.1-2. Connection with the Bessel functions and the modified Bessel functions:

    Ai(

    ) =1

    3

    1

    3(

    )

    1

    3(

    )

    =

    1 1

    3

    1

    3(

    ),

    =

    2

    3

    3 2

    ,Ai(

    ) = 13

    1 3( ) +

    1 3( ) ,

    Bi(

    ) = 13

    1 3( ) +

    1 3( ) ,

    Bi(

    ) = 13

    1 3( )

    1 3( ) .

    A.8.2. Power Series and Asymptotic Expansions

    A.8.2-1. Power series expansions as

    0:

    Ai(

    ) = 1 (

    ) 2 (

    ),

    Bi(

    ) =

    3

    1 (

    ) + 2 (

    ) ,

    (

    ) = 1 +1

    3!

    3 +1 4

    6!

    6 +1 4 7

    9!

    9 +-

    - -

    =

    =0

    3

    1

    3

    3

    (3

    )!,

    (

    ) =

    +2

    4!

    4 +2 5

    7!

    7 +2 5 8

    10!

    10 +-

    - -

    =

    =0

    3

    2

    3

    3

    +1

    (3

    + 1)!,

    where 1 = 32

    3

    (2 3) 0.3550 and 2 = 31

    3

    (1 3) 0.2588.

    A.8.2-2. Asymptotic expansions as

    .

    For large values of

    , the leading terms of asymptotic expansions of the Airy functions are

    Ai(

    ) 12

    1 2

    1 4 exp( ), = 23

    3 2,

    Ai(

    )

    1 2

    1 4 sin

    +

    4

    ,

    Bi(

    )

    1 2

    1 4 exp(

    ),

    Bi(

    )

    1 2

    1 4

    cos

    +

    4

    .

    Reference: M. Abramowitz and I. Stegun (1964).

    2002 by Chapman & Hall/CRC

  • 7/28/2019 c2999 PDF Supa

    15/26

  • 7/28/2019 c2999 PDF Supa

    16/26

    A.9.1-2. Kummer transformation and linear relations.

    Kummer transformation:

    ( , ;

    ) =

    ( , ;

    ), ( , ;

    ) = 1

    (1 + , 2 ;

    ).

    Linear relations for

    :

    ( ) ( 1, ;

    ) + (2 +

    ) ( , ;

    )

    ( + 1, ;

    ) = 0,

    ( 1) ( , 1;

    ) ( 1 +

    ) ( , ;

    ) + ( )

    ( , + 1;

    ) = 0,

    (

    + 1)

    (

    ,

    ;

    ) (

    + 1,

    ;

    ) + (

    1)

    (

    ,

    1;

    ) = 0,

    (

    ,

    ;

    )

    (

    1,

    ;

    )

    (

    ,

    + 1;

    ) = 0,

    (

    +

    )

    (

    ,

    ;

    ) (

    )

    (

    ,

    + 1;

    )

    (

    + 1,

    ;

    ) = 0,

    ( 1 +

    ) ( , ;

    ) + ( ) ( 1, ;

    ) ( 1) ( , 1;

    ) = 0.

    Linear relations for

    :

    ( 1, ;

    ) (2 +

    ) ( , ;

    ) + ( + 1) ( + 1, ;

    ) = 0,

    ( 1) ( , 1;

    ) ( 1 +

    ) ( , ;

    ) +

    ( , + 1;

    ) = 0,

    ( , ;

    )

    ( + 1, ;

    ) ( , 1;

    ) = 0,

    ( ) ( , ;

    )

    ( , + 1;

    ) + ( 1, ;

    ) = 0,

    (

    +

    )

    (

    ,

    ;

    ) +

    (

    1)

    (

    + 1,

    ;

    )

    (

    ,

    + 1;

    ) = 0,

    (

    1 +

    )

    (

    ,

    ;

    )

    (

    1,

    ;

    ) + (

    + 1)

    (

    ,

    1;

    ) = 0.

    A.9.1-3. Differentiation formulas and Wronskian.

    Differentiation formulas:

    ( , ;

    ) =

    ( + 1, + 1;

    ),

    ( , ;

    ) =

    ( + 1, + 1;

    ),

    ( , ;

    ) =( )

    ( )

    ( + , + ;

    ),

    ( , ;

    ) = (1) ( )

    ( + , + ;

    ).

    Wronskian:

    ( , ) =

    =

    ( ) ( )

    .

    A.9.1-4. Degenerate hypergeometric functions for

    = 0, 1, 2,

    :

    ( , + 1;

    ) =(1) 1

    ! ( ) ( , +1;

    ) ln

    +

    =0

    ( )

    ( + 1)

    ( + o )

    (1 + o )

    (1 + + o )

    o ! +

    ( 1)! ( )

    1

    =0

    ( )

    (1 )

    o !,

    where

    = 0, 1, 2,

    (the last sum is dropped for

    = 0),

    (

    ) = [ln (

    )] is the logarithmic

    derivative of the gamma function,

    (1) =

    ,

    (

    ) =

    +

    1

    =1

    1,

    where

    = 0.5772

    is the Euler constant.

    2002 by Chapman & Hall/CRC

  • 7/28/2019 c2999 PDF Supa

    17/26

    If < 0, then the formula

    (

    ,

    ;

    ) = 1

    (

    + 1, 2

    ;

    )

    is valid for any

    .

    For

    0, 1, 2, 3,

    , the general solution of the degenerate hypergeometric equation can

    be represented in the form =

    1

    ( , ;

    ) + 2

    ( , ;

    ),

    and for

    = 0, 1, 2, 3,

    , in the form

    = 1

    1

    (

    + 1, 2

    ;

    ) + 2

    (

    + 1, 2

    ;

    )

    .

    A.9.2. Integral Representations and Asymptotic Expansions

    A.9.2-1. Integral representations:

    ( , ;

    ) = ( )

    (

    )

    (

    )

    1

    0

    1(1 )

    1

    (for > > 0),

    (

    ,

    ;

    ) =1

    (

    )

    0

    1(1 +

    )

    1

    (for

    > 0,

    > 0),

    where ( ) is the gamma function.

    A.9.2-2. Integrals with degenerate hypergeometric functions:

    ( , ;

    )

    = 1

    1

    ( 1, 1;

    ) +

    ,

    ( , ;

    )

    =1

    1

    ( 1, 1;

    ) +

    ,

    ( , ;

    )

    = !

    +1

    =1

    (1)

    +1(1 )

    +1

    (1 )

    (

    + 1)!

    (

    ,

    ;

    ) +

    ,

    ( , ;

    )

    = !

    +1

    =1

    (1)

    +1

    +1

    (1 )

    (

    + 1)!

    (

    ,

    ;

    ) +

    .

    A.9.2-3. Asymptotic expansion as |

    | :

    ( , ;

    ) = (

    )

    ( )

    =0

    (

    )

    (1

    )

    !

    + !

    ,

    > 0,

    ( , ;

    ) ="

    ( )

    "

    ( )(

    )

    #

    =0

    ($

    )

    ($

    + 1)

    % !(

    &) +

    !,

    &< 0,

    ($

    , ;&

    ) =&

    #

    =0

    (1) (

    $)

    ($

    + 1)

    % !&

    + !

    ,