500 INTEGRALS

OF
ELEMENTARY AND SPECIAL
FUNCTIONS
⊕
Francis J. O’Brien, Jr.

GAMMA FUNCTION
500 Integrals of Elementary and
Special Functions
Library of Congress Cataloging-in-Publication Data

O’Brien, Francis Joseph, Jr.

500 Integrals of Elementary and Special Functions
p. cm.
Includes bibliographical references and index.
ISBN: 1-4392-1981-8
ISBN-13: 978-1439219812
1. Mathematics. 2. Differential and integral calculus.
3. Proofs and derivations.
Library of Congress Catalog Card Number:

Copyright © 2008 by Francis Joseph O’Brien, Jr. All rights reserved. No part of this publication
may be reproduced, stored in retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording, or otherwise, without the written permission of
the author, except as permitted under Sections 107 or 108 of the 1976 United States Copyright
Act.

Printed in the United States of America

500 Integrals of Elementary
and Special Functions
Francis J. O’Brien, Jr.
Naval Undersea Warfare Center
Naval Sea Systems Command
Newport, Rhode Island
For my daughter, Miss Lily-Rae O’Brien

⊗
In gratitude to The Department of Defense
 PREFACE
▲
●
▼
This book is a listing of solved formulas for about 500 integrals, sums, series, and
products. The intended audience is students and practitioners in the fields of mathematics, pure
and applied science, and engineering. The fundamental purpose of the book is to present a
modest listing of formulas with worked out solutions structured on the widely used desk
reference, Gradshteyn & Ryzhik’s Table of Integrals, Series, and Products (2007).
We provide “new” formulas in a limited number of areas—exponential and logarithmic
functions and selected special functions with emphasis on the gamma and related functions. The
level of difficulty of the material in this book ranges from easy to moderately difficult, and
covers selected topics in first year calculus through advanced calculus. The focus is on
understanding how to evaluate the unsolved indefinite and definite integrals covered in the text.
Throughout the computer is used as an answer- checker rather than the primary evaluator or
prover.
In this age of computer technology, one may wonder why books such as Table of
Integrals, Series, and Products are even necessary. One of the many sophisticated computer
programs used in the scientific, engineering and mathematical fields is Mathematica
(http://www.wolfram.com/). This computer algebra system can display solutions to derivatives
& integrals in a matter of seconds. It has been asserted that Mathematica can solve
approximately 80% of the formulas in Table of Integrals, Series, and Products. Yet the
Gradshteyn & Ryzhik book is still widely purchased, and is now available in the 7th edition.
Computer calculators such as Mathematica can solve the majority of integrals in
Gradshteyn & Ryzhik and this book. In addition, comprehensive mathematics websites such as
The Wolfram Functions Site at http://functions.wolfram.com/ provide thousands of formulas, yet
very little exists in the way of proofs, derivations, and justifications. The only significant effort I
am aware of is a series of journal articles by Prof. V.H. Moll of Tulane University. He is in the
process of deriving and verifying the formulas in Gradshteyn & Ryzhik. See the website,
http://www.math.tulane.edu/~vhm/.
Recent papers that amplify the derivations given in this book may be found at the
docstoc.com website. They derive relations for the Pochhammer Symbol, double factorials, and
gamma transformations, identities and special values. See
http://www.docstoc.com/profile/waabu.
Given the significant decline in enrollments in mathematics, science and engineering in
the United States, it is important to supplement major mathematical reference works such as the
Gradshteyn & Ryzhik handbook. Other writers may be motivated to assist in this effort.
 Structure of the Book
As mentioned, 500 Integrals is based on the architecture of Gradshteyn & Ryzhik’s Table
of Integrals, Series, and Products, 7th Edition (GR for short). In essence, the book contains
previously published formulas in GR as well as candidate formulas for the next edition. The
material is presented in the order of appearance in GR using their section nomenclature and
formula numbering system. This format is used instead of conventional chapter and sub-chapter
headings.
The topics covered reflect the author’s own interests in theoretical fundamentals and
potential applications in education, science and engineering. The first section, Mathematical and
Graphical Summary of Selected Elementary and Special Functions, contains an abbreviated
tutorial on all the notations, definitions and properties of the functions used in this book. This
material is based on various standard sources including GR (Sections 0. and 1.) and new results.
Key references are provided for additional background reading and study. The Math. Summary
should be consulted for guidance in the solution of the formulas presented. The Notations
section also contains useful information. The first major section of results consists of Indefinite
Integrals of selected Elementary Functions (Sections 2.3 & 2.7 in GR). This is followed by
Definite Integrals of Elementary Functions (Sections 3.3–3.4 & 4.2.–4.4 in GR), and lastly,
selected Special Functions (Sections 8.2 & 8.3 in GR).
A number of integrals are left as exercises. In addition, certain related integrals are
solved “from scratch” so that readers can search for a simplified solution based on earlier
material (or their own creation) while other calculations are stated as “similar to above” without
solution, inviting readers to supply details.
Speed of identification of a formula is a major concern for many users of mathematical
reference handbooks. At the rear of the book is an Index of Formulas which lists all of the
formulas presented in the book, arranged by section and page numbers corresponding to the
Table of Contents. An Index of Symbols, Functions and Concepts may also be found in the rear
of the book.
In the author’s opinion the most useful formulas presented in this volume are those
involving three-parameter algebraic-exponential functions expressed in terms of gamma
functions (see Mathematical Summary [Incomplete Gamma Function—Indefinite Integrals], and
Sections 2.32, 3.326, 3.381, 3.462). I view them as “reduction” formulas which help solve and
create a number of useful integrals.
Some formulas developed here have been used in the derivation of new elementary
probability models; see “Summary of Four Generalized Exponential Models (GEM) For
Continuous Probability Distributions,” Jan. 18, 2008, arXiv:0801.2941v2 [math.GM].

Notes on the entries in the book—

• All integrals in this book omit the constant of integration

• In most cases, derived solutions may be verified by ordinary differentiation or by
differentiation under the integral sign using Leibnitz’s rules provided in
Mathematical Summary
o Computer verification is suggested when this is possible
• Variables are considered to be real quantities, unless otherwise indicated
 1
• A square root x or x 2 or π π is taken to be positive unless otherwise
specified
• Logarithms are natural logarithms, denoted ln x (base e), unless otherwise
indicated
• Errata for the 6th and 7th editions of Gradshteyn and Rhyzik can be obtained
online at http://www.mathtable.com/gr/
• “GR” or “G & R 7e” refer to Gradshteyn and Rhyzik’s Table of Integrals, Series
and Products, 7th edition, unless otherwise indicated
• math.com refers to the public Internet website, http://www.quickmath.com/,
which calculates indefinite and definite one-dimensional integrals, and performs
other services including derivatives, partial fraction expansion, graphical plotting,
matrix inversion, etc.
o math.com displays the solutions for the lower incomplete gamma function
as Γ(a ) − Γ(a, x ) vice γ (a, x )
• The Mathematica Integrator calculates indefinite integrals at
http://integrals.wolfram.com/index.jsp
o Sometimes math.com unable but Mathematica able to do an indefinite integral
o Sometimes an indefinite integral can be calculated while the definite one
cannot
o Sometimes computer provides a solution at higher level of complexity than
dγ (a, x )
needed; e.g., the derivative, (see Section 8.356)
da
o Sometimes computer returns a “different” solution compared to paper and
pencil answer
¾ Computer always provides only a single solution when multiple
ln (a + bx )dx e ax − 1
answers are possible; e.g., ∫ or ∫ ax dx (see Section
x e +1
2.32)
• “verified on math.com”, “verified on mathematica” means author’s answer
confirmed by computer when possible
o Post-solution simplification is often required for final form
o Logarithmic indefinite integrals: computer calculation usually returns an
incomplete gamma function solution vice desired reduction formula by
integration by parts/change of variable or combination solution
• “not verifiable on math.com” & “not verifiable on mathematica” means computer
unable to calculate answer directly from the input, providing evidence that the
human calculator is still useful in this electronic age.

———

I would appreciate communications regarding misprints and any helpful suggestions for
improvement of presentation of the material. I can be reached by e-mail at either
francis.j.obrien@navy.mil —or— francis28@cox.net.
 Acknowledgements
The author would like to acknowledge and thank all those who have funded, assisted, and
encouraged him over the years. These include the Office of Naval Research, the Base
Commander, my research colleagues in the USW Combat Systems Department, and the
attorneys and paralegals in Office of Patent Counsel of the Naval Undersea Warfare Center,
Newport, Rhode Island.
Professor Alan Jeffrey of the University of Newcastle Upon Tyne (England), Editor of
Table of Integrals, Series and Products, has been most helpful in his encouragement. I also thank
his co-editor, Dr. Daniel Zwillinger of Rensselaer Polytechnic Institute, for correspondence
regarding errata and related matters.
I also want to acknowledge the assistance of Aimee Ross, a 3rd year mathematics major at
University of Massachusetts—Dartmouth, who read the early sections of the manuscript, and
provided feedback on the level of difficulty and the clarity of the material.

Francis J. O’Brien, Jr.

Newport, Rhode Island
October 12, 2008
 TABLE OF CONTENTS
Section in Gradshteyn and Ryzhik Page
MATHEMATICAL AND GRAPHICAL SUMMARY OF SELECTED ELEMENTARY
AND SPECIAL FUNCTIONS

Exponential Functions 2
Logarithmic Functions 4
Gamma Function 6
Incomplete Gamma Functions 8
Probability Integral or Error Function (Erf) and Imaginary Error Function (Erfi) 10
Exponential-Integral Function 12
Logarithm-Integral Function 14
Euler’s Constant 16
Catalan’s Constant 17
Partial Fractions 18
Miscellaneous (Completing the Square, Finite Binomial Expansions, Double 20
Factorial, Natural Number N, Differentiation Under the Integral Sign)

NOTATIONS 22

INTRODUCTION
0.1 Finite Sums 24

0.11 Progressions 25
0.111 Arithmetic progressions 25
0.112 Geometric progressions 26
0.12 Sums of powers of natural numbers 26

INDEFINITE INTEGRALS OF
ELEMENTARY FUNCTIONS

2.3 The Exponential Function 28

n
2.31 Forms containing e ax , e ax 30
2.312 31
2.32 The exponential combined with rational functions of x 35

2.7 Logarithms and Inverse and Hyperbolic Functions 61

2.71 The logarithm 62

 2.72-2.73 The logarithm and combinations of logarithms and algebraic functions 62

DEFINITE INTEGRALS OF
ELEMENTARY FUNCTIONS

3.3–3.4 Exponential Functions 79

3.31 Exponential functions 80

3.310 81
3.311 82
3.32–3.34 Exponentials of more complicated arguments 84
3.321 85
3.322 89
3.323 90
3.324 94
3.326 96
3.327–3.334 Exponentials of exponentials 98
3.327 99
3.328 100
3.331 101
3.35 Combinations of exponentials and rational functions 103
3.351 104
3.353 105
3.36–3.37 Combinations of exponentials and algebraic functions 106
3.361 107
3.362 108
3.363 109
3.371 116
3.38–3.39 Combinations of exponentials and arbitrary powers 120
3.381 121
3.382 125
3.41–3.44 Combinations of rational functions of powers and exponentials 126
3.427 127
3.434 128
3.46–3.48 Combinations of exponentials of more complicated arguments and powers 129
3.461 130
3.462 134
3.464 148
3.471 151
3.473 160

4.2–4.4 Logarithmic Functions 161

4.21 Logarithmic functions 162

4.211 163
 4.212 168
4.215 174
4.22 Logarithms of more complicated arguments 177
4.229 178
4.24 Combinations of logarithms and algebraic functions 179
4.241 180
4.26–4.27 Combinations involving powers of the logarithm and other powers 181
4.269 182
4.272 185
4.274 190
4.28 Combinations of rational functions of ln x and powers 192
4.281 193
4.283 194
4.29-4.32 Combinations of logarithmic functions of more complicated arguments 195
and powers
4.326 196
4.33–4.34 Combinations of logarithms and exponentials 197
4.331 198
4.337 199

SPECIAL FUNCTIONS
8.2 The Exponential Integral Function and Functions Generated by It 202

8.21 The exponential integral function Ei(x) 203

8.212 204
8.24 The logarithm integral li(x) 210
8.240 211
8.241 Integral representations 212
8.25 The probability integral, the Fresnel Integrals Φ ( x ), S ( x), C ( x) , the error 214
function erf(x), and the complementary error function erfc(x)
8.250 Definition 215
8.252 Integral representations 216

8.3 Euler’s Integrals of the First and Second Kinds and Functions Generated by Them 221

8.31 The gamma function (Euler’s integral of the second kind): Γ(z) 222
8.313 223
8.33 Functional relations involving the gamma function 224
8.331 225
8.334 232
8.335 238
8.339 Particular Values: For n a natural number 239
8.35 The Incomplete Gamma Function 253
8.350 Definition 254
8.351 255
8.352 Special cases 256
 8.353 Integral representations 261
8.356 Functional relations 263
8.359 Relationships with other functions 265
8.36 The Psi function ψ ( x ) 273
8.367 Euler’s constant: Integral representations 274
References 276

Index 278
Index of Formulas 284

LIST OF ILLUSTRATIONS

Figure 1. Exponential Function (Growth) 2

Figure 2. Exponential Function (Decay) 4
Figure 3. Natural Logarithm Function 4
Figure 4. Gamma Function 6
Figure 5. Error Function 10
Figure 6. Imaginary Error Function 10
Figure 7. Exponential-integral Function 12
Figure 8. Logarithm-integral Function 14
Figure 9. Euler’s Constant 16
Figure 10. Catalan’s Constant 17
MATHEMATICAL & GRAPHICAL SUMMARY
OF SELECTED
ELEMENTARY AND SPECIAL FUNCTIONS

GAMMA FUNCTION
 2

EXPONENTIAL FUNCTIONS

e− x
x
e

x x

Figure 1. Exponential Function (Growth) Figure 2. Exponential Function (Decay)

Notation: e x = exp(x ) = ln −1 (x )

Definitions & Laws of Exponents:

a 0 = 1, a ≠ 0 a x = e x ln a x = e ln x
1

( a)
p

[( )]
q p
a = a = a =
q p q p q

a x+ y = a x a y a − x = e − x ln a 1 1 ( ) =e
x n = e ln x
n n ln x

= [(− a ) ] = [(− 1) ] [a ]
1

(− a ) q = (e ) = e
p q q q
ax = ex
n n
ax ln a p p p a
ln x n
a
n a ln x
a x− y = xn
ay a −x = e−x
n n
ln a 1

(a ) (− 1) = i (imaginary)
2
x y
=a xy

(ab )x = a xb x
64 4 4 4 74 4 4 4 8
Let a be e for exp.

Limits:

lim e x = 1 lim e x = +∞ ex e − ax − e − bx
x →0 x → +∞ lim a = +∞ lim =b−a
x → +∞ x x →0 x
−x
lim e =1 lim e − x = 0
x →0 x → +∞
e−x xa e −ax − e −bx
lim a = lim x = 0 lim =0
x → +∞ x x → +∞ e x →∞ x
 3

Derivatives:
¾ a, e, n are assumed constants except where noted

du v de v ln u
n n
dv ln u du dv da x de x ln a de x
= = uv = vu v −1 + u v ln u = = nx n −1a x ln a = ex
n

dx dx dx dx dx dx dx dx
da u
de u ln a
du x x ln a de − x
= = a u ln a da
=
de
= a x ln a = −e − x
dx dx dx dx dx dx
u de ax
de
= eu
du
d x
1 = ae ax
− x ln a dx
dx dx a = de =− x
ln a
de − ax
6444444447444444448 dx dx a = −ae − ax
chain rule n n dx
da x de x ln a
= = a x x n ln x ln a
n

dn dn
[n not a constant ]

Indefinite Integrals:
−x −x
∫ e dx = e ∫ e dx = −e
x x

e ax − ax e − ax
∫ e dx = ∫ e dx = −
ax

a a

∫x
± m ± ax n
• See Section 2.32 for generalized indefinite exponential integrals of form, e dx, expressed in
terms of incomplete gamma functions. Summarized in INCOMPLETE GAMMA FUNCTIONS.
• These exponential integrals are used often in the evaluation of definite integrals of elementary and some
special functions.

References
• Bers, Calculus. Ch. 6, pp. 367 ff. & 375 ff ; Ch. 7, pp. 453 ff.; Ch. 8, pp. 547 ff.
• Carr, 1970
• Dwight, Table of Integrals, 1961
• Spiegel, Chaps. 7 & 20
• Table of Integrals, Series, and Products. Sects. 0.245, 1.2 (“The Exponential Function”), 2.01 (“The basic
integrals”)
 4

LOGARITHMIC FUNCTIONS
(Base e)

1
ln x

1
x

1 1
Figure 3. Natural Logarithm Functions, ln x & ,&
ln x x

Notation:

log x = ln x (natural logarithm, base e)

log m x = (log x )
m

x
dt
Definition: ∫
1
t
= log x, x > 0
 5

Properties:

log ( xy ) = log x + log y log x n = n log x ln e x = x

x
log = log x − log y log
1
= − log x ln e − x = − x
y x ln e ax = ax
⎛ b⎞
( )
a ln x + b ln y = ln x a y b = a ln⎜⎜ xy a ⎟⎟ ln e − ax = − ax
⎝ ⎠
⎛ ⎞
⎛ xa ⎞ ⎜ x ⎟
a ln x − b ln y = ln⎜⎜ b ⎟⎟ = a ln⎜ b ⎟
⎝y ⎠ ⎜ ya ⎟
⎝ ⎠
Limits:

lim log x = +∞ xa log x ex ln(1) = 0

x → +∞ lim = +∞ lim =0 lim = +∞
lim log x = −∞
x → +∞ log x x → +∞ e x x → +∞ log x ln (e ) = 1
x →0 + log x
log x lim x = −∞
x
e ⎛1⎞
lim a = 0 x →0 + e lim =0 ln⎜ ⎟ = −1
x → +∞ x x →0 + log x ⎝e⎠

Derivatives:

d ln x 1 d ln m (a + bx ) mb ln m−1 (a + bx ) d ln u 1 du
= (chain rule)
= ,x ≠ 0 =
dx x dx a + bx dx u dx

Indefinite Integrals:

dx
∫ x
= ln x

NOTE: see Section 2.32 for special logarithmic function—the dilogarithm,

ln (1 − x )
Li 2 ( x ) = − ∫ dx
x
∞
xk
Li 2 ( x ) = ∑ 2 , x < 1
k =1 k

References
• Bers, Calculus. Ch. 6, pp. 357 ff.; Ch. 7, pp. 454 ff.; Ch. 8, pp. 547 ff.
• Carr, 1970
• Dwight, Table of Integrals, 1961
• Spiegel, Chaps. 7 & 20
• Table of Integrals, Series, and Products. Sects. 1.5 (“The logarithm”), 2.01 (“The basic integrals”)
• Wolfram Functions website, http://functions.wolfram.com/GammaBetaErf/
 6

GAMMA FUNCTION

Figure 4. Gamma Function, Γ( x )

Definition:
∞
Γ(z ) = ∫ t z −1e −t dt , real z > 0 [Formula 8.310.1]
0
z −1
⎛1⎞
1
Γ(z ) = ∫ ln⎜ ⎟ dt , real z > 0 [Formula 8.312.1]
0 ⎝t ⎠

Integral, Product, and Series Representations:

∞
Γ( x ) = z x
∫t
x −1 − zt
e dt , real z , x > 0 [Formula 8.312.2]
0

p! p x px
Γ( x ) = lim = lim , real x > 0 [Artin, Formula 2.7]
⎛ x⎞
p →∞ p p →∞ p
x∏ ( x + k ) x ∏ ⎜1 + ⎟
k =1 k =1 ⎝ k⎠
x
⎛ x⎞ ⎛ 1⎞
exp⎜ ⎟ ∞ ⎜
1+ ⎟
exp(− γx ) ∞
⎝ k⎠ 1 ⎝ k⎠ nx n
Γ(x ) =
k
x
∏ x
= ∏
x k =1 x
= lim
n →∞ x
∏ k + x, real x > 0 [Formula 8.322]
k =1
1+ 1+ k =1
k k
n −1
⎛ ⎞
Γ⎜ z + ⎟ & Γ(nz ) &
k 1
∏
k =0 ⎝ n⎠ Γ( x )
[Section 8.335]
⎧ ∞
⎡ ⎛ x ⎞ x ⎤⎫
log Γ( x ) = − ⎨ln x + γx + ∑ ⎢ln⎜1 + ⎟ − ⎥ ⎬, real x > 0 [Artin, Formula 2.9]
⎩ k =1 ⎣ ⎝ k ⎠ k ⎦⎭

Properties:

Γ( x ) = (x − 1)! Γ(1 − x ) lim Γ( x ) = +∞ lim Γ( x ) = +∞

Γ(− x ) = − , x not a natural x →0 + x → +∞
x
num.
Γ( x ) ≠ 0 Γ( x + 1) = xΓ( x ) lim Γ( x ) = −∞ lim Γ( x ) is indeterminate
x →0 − x → −∞
 7

Derivatives:

dΓ( x )
∞ ∞
= Γ ′( x ) = Γ′(1) = ∫ e −t ln tdt
d x −1 −t
dx ∫0
t e dt =
dx 0
∞
∂
∞
= −γ (Euler' s constant )
∫ ∂x e
( x −1) ln t
e −t dt = ∫ t x −1e −t ln tdt = Γ( x )Ψ ( x )
0 0
= Ψ (1) (Formula 8.366.1)
∞ d ln Γ(x ) Γ′(x )
Γ (n ) ( x ) = ∫ t x −1e −t (ln t ) dt
n • = = Ψ (x )
dx Γ( x )
0
(digamma or Psi function), Formula 8.330
d ⎛ 1 ⎞ 1 dΓ( x ) Ψ (x )
• ⎜⎜ ⎟⎟ = − =−
dx ⎝ Γ( x ) ⎠ [Γ(x )] dx
2
Γ( x )

Relation to Incomplete Gamma Functions, γ (a, x ) & Γ(a, x ) :

Γ(a ) = γ (a, x ) + Γ(a, x )

∞ x ∞

∫t = ∫ t e dt + ∫ t a −1e −t dt
a −1 −t a −1 −t
e dt [Formula 8.356.3]
1424
0
3 1
0
424 3 1x
424 3
COMPLETE GAMMA 1 4 4
LOWER
42 4 4 4
UPPER
3
INCOMPLETE GAMMA
• NOTE: Relation is additive; e.g., γ (a, x ) = Γ(a ) − Γ(a, x )

Trigonometric Functions:
• Bers, Ch. 6.
• Dwight, Table of Integrals, 1961
• Spiegel, Ch. 5.
• Table of Integrals, Series, and Products, Sects. 1.3-1-4, 8.31-8.35 & Index

NOTE on Γ(x ) : Emil Artin’s brief 1964 book—The Gamma Function—provides a complete statement for
real variables of this special transcendental function called the complete gamma function, and derives the
fundamental mathematical properties originated in the classical 18th & 19th century works of Euler, Gauss,
Legendre, Riemann, Stirling, Weierstrass, and others. See Artin’s book for other definitions, theorems, and
derivations of Γ(x ) not given in this elementary book. Whittaker & Watson is also recommended for
derivations.

References
• Abramowitz & Stegun, Ch. 6.
• Artin, 1964
• Bers, Calculus , Chaps. 4 & 6, pp. 402-3
• Carr, 1970
• Moll, http://www.math.tulane.edu/~vhm/Table.html
• Table of Integrals, Series, and Products, Sects 1.3-1.4 and Sects. 8.31-8.35 & Index
• Whittaker & Watson, 1934
• Wolfram Functions website, http://functions.wolfram.com/GammaBetaErf/
• O’Brien, http://www.docstoc.com/docs/5473276/Pochhammer-Symbol-Selected-Proofs
• O’Brien, http://www.docstoc.com/docs/5836783/Selected-Transformations-Identities--and-Special-
Values--for-the-Gamma-Function
 8

INCOMPLETE GAMMA FUNCTIONS

1. Lower Incomplete Gamma Function

Definition: γ (a, x ) = ∫ t a −1e −t dt , real a > 0 [Formula 8.350.1]

x

NOTE:
• See Sect. 8.352 for integer special cases, γ (n, x ) , etc.
1
Integral Representation: γ (a, x ) = x a ∫ t a −1e − xt dt , real a,x > 0 [Sections 3.331 & 8.353]
0

Properties:

• γ (a, x ) = Γ(a ) − Γ(a, x )

γ (a + 1, x ) + x a e − x
• γ (a, x ) = , a ≠ 0 [Formula 8.356.1]
a
• γ( 1
, x ) = π Φ(x ) ,
2
[ Φ(x ) = erf(x), error function, 8.250.1]
( )
2

• γ ( , x) = π Φ x
1
2

• γ (a,0) = 0 [Formula 8.350.5]

• γ (a, ∞ ) = Γ(a )
dγ (a, x ) dΓ(a, x ) a −1 − x
• =− = x e [Form. 8.356.4]
dx dx
dγ (a, u ) dΓ(a, u ) dγ (a, u ) du
• =− = (chain rule)
dx dx du dx
dγ (a, x )
1
• = γ (a, x ) ln x + x a ∫ t a −1e − xt ln tdt [Sect. 8.356]
da 0

2. Upper Incomplete Gamma Function

∞
Definition: Γ(a, x ) = ∫ t a −1e −t dt [Formula 8.350.2]
x

NOTE:
• See Sect. 8.352 for integer special cases, Γ(n, x ) , etc.

∞
Integral Representation: Γ(a, x ) = x a
∫t
a −1 − xt
e dt , real a,x > 0 [Sections 3.331 & 8.353]
1

Properties:

• Γ(a, x ) = Γ(a ) − γ (a, x )

Γ(a + 1, x ) − x a e − x
• Γ(a, x ) = , a ≠ 0 [Formula 8.356.2]
a
• Γ(12 , x 2 ) = π − π Φ ( x ) [ Φ(x ) = erf(x), error function, 8.250.1]
• Γ( 12 , x ) = π − π Φ ( x)
 9

• Γ(a,0 ) = Γ(a )[Formula 8.350.3]

• Γ(a, ∞ ) = 0 [Formula 8.350.4]
dΓ(a, x ) dΓ(a, u )
• & (see 1. above)
dx dx
dΓ(a, x )
∞
• = Γ(a, x ) ln x + x ∫ t a −1e − xt ln tdt
a
[Sect. 8.356]
da 1

d ⎡ γ (a, x ) ⎤ γ (a, x ) x a 1
[ ( )]
Γ(a ) Γ(a ) ∫0
= − Ψ + t a −1e − xt ln tdt , real a,x > 0 [Section 8.356]
da ⎢⎣ Γ(a ) ⎥⎦
ln x a

NOTE:
• See exponential-integral function, Ei( x) , for relation to Γ(a, x )

Indefinite Integrals:

(− 1)1−γ Γ(γ ,−ax n ) = (− 1)1−γ ∞

m +1
∫ x e dx = ∫t
γ −1 −t
e dt , γ =
n
m ax
γ γ
na na − ax n
n
(
Γ γ , βx n 1 ) ∞
m +1
∫ x e dx = − ∫t
m − βx γ −1 −t
=− γ e dt , γ =
n

γ
[Formula 2.33.10]
nβ nβ βx n
n
e ax
n
(− 1) z +1
(
a z Γ − z ,− ax n ) (− 1)z +1 a z ∞
1 m −1
∫ x m dx = n
=
n ∫ t z +1
e −t dt , z=
n
[Formula 2.325.6]
− ax n

e − βx
n
β z Γ − z , βx n (
βz ) ∞
e −t m −1
∫ xm dx = −
n
= −
n ∫n t z +1 dt , z=
n
[Formula 2.33.19]
βx

NOTE:
• These elementary algebraic-exponential formulas are used extensively in this book to derive the definite integral
expressions (see Formulas—3.326.2, 3.462.19, 3.381.8-3.381.10 & others in those sections—for negative
exponential forms), to prove existing formulas, and to create new integrals for a useful class of transcendental
functions and special functions.
⎛ 1 ⎞ ⎛1 ⎞
¾ See Section 8.359 for special forms, Γ⎜ ± ,± x⎟ &γ ⎜ , ± x⎟
⎝ 2 ⎠ ⎝2 ⎠

References
• Abramowitz & Stegun, Ch. 6
• Boas, Ch. 11
• Moll, http://www.math.tulane.edu/~vhm/Table.html
• Table of Integrals, Series, and Products, Use of the Tables (“The factorial gamma”), Sect. 8.35
• Wolfram Functions website, http://functions.wolfram.com/GammaBetaErf/Gamma2/
 10

PROBABILITY INTEGRAL OR ERROR FUNCTION (ERF)

AND
IMAGINARY ERROR FUNCTION (ERFI)

x
x

Figure 5. Error Function (erf) Figure 6. Imaginary Error Function (erfi)

Definition:
x
Φ (x ) = erf (x ) =
2
π ∫
−t 2
e dt [Formula 8.250.1]
0
∞
erfc(x ) = 1 − Φ (x ) =
2
∫ e dt [Complimentary Error Function, Formula 8.250.4]
−t 2

π x

Integral Representation:
x2
e −t
Φ (x ) = erf (x ) =
1
π ∫0 t
dt [Formula 8.251.1]

Properties:

Φ (− x ) = −Φ ( x )
Φ (0 ) = 0 Sect. 8.359

Φ (± ∞ ) = ±1
 11

Relation to Incomplete Gamma Function:

⎛1 ⎞ ⎛1 ⎞
γ ⎜ , x2 ⎟ γ ⎜ , x⎟
Φ(x ) = ⎝
2 ⎠ Φ ( ) x = ⎝
2
π
⎠
π

Imaginary Error Function, erfi(z):

erf(iz )
erfi( z ) =
i
(iz )2
e −t
z iz
2 2 1
∫e ∫e ∫
−t 2
erfi( z ) = t2
dt = dt = dt
π 0 i π 0 i π 0 t

Relation to Incomplete Gamma Function:

⎛1 ⎞
γ ⎜ ,− z 2 ⎟
erfi(z ) = ⎝ ⎠
2
i π

Indefinite Integrals:

π
∫e
−t 2
dt = erf (t )
2
π
∫e
t2
dt = erfi(t )
2

References
• Abramowitz & Stegun, Ch. 7.
• Boas, Ch. 11.
• Dwight, Table of Integrals, 1961
• Moll, http://www.math.tulane.edu/~vhm/Table.html
• Spiegel, Ch. 35
• Table of Integrals, Series, and Products, Use of the Tables (“Exponential and related functions”), and
Sect. 8.250.
• Wolfram Functions website, http://functions.wolfram.com/GammaBetaErf/Erf2/
 12

EXPONENTIAL-INTEGRAL FUNCTION

Ei ( x )

e−x
Figure 7. Exponential-integral Function, Ei(x), −
x

Definition:

Formulas 8.211.1 & 8.211.2

∞
e −t et
x
⎡−ε e −t ∞ −t
⎤
Ei( x ) = − ∫ dt = ∫ dt , x < 0 Ei( x ) = − lim ⎢ ∫ dt + ∫
e
dt ⎥, x > 0 (Cauchy Principal Value PV)
t t ε → +0
−x −∞ ⎣− x t ε t ⎦

Properties:

Ei(+ ∞ ) = +∞ Ei(− ∞ ) = 0 Ei(0 ) is not defined

Relation to Logarithm-Integral Function:

( )
Ei( x ) = li e x , x < 0

( )
Ei(ax ) = li e ax , x < 0, a ≠ 0
 13

Relation to Incomplete Gamma Function:

∞ ∞
e −t e −t
Ei( x ) = − ∫ dt = −Γ(0,− x ) Ei(− x ) = − ∫ dt = −Γ(0, x )
−x
t x
t

∞ ∞
e −t e −t
− Ei( x ) = ∫ dt = Γ(0,− x ) − Ei(− x ) = ∫ dt = Γ(0, x ) Form. 8.359.1
−x
t x
t

Indefinite Integrals:

e±s
∫ s ds = Ei(± s )

References
• Abramowitz & Stegun, Ch. 5.
• Moll, http://www.math.tulane.edu/~vhm/Table.html
• Table of Integrals, Series, and Products, Use of the Tables (“Exponential and related functions”), and
Sect. 8.21 & Sects. 3.04-3.05 (“Improper Integrals”; “Principal Values”)
• Weisstein, Eric W. "Exponential Integral." From MathWorld--A Wolfram Web Resource.
http://mathworld.wolfram.com/ExponentialIntegral.html
• Wolfram Functions website, http://functions.wolfram.com/GammaBetaErf/ExpIntegralE/
 14

LOGARITHM-INTEGRAL FUNCTION

Figure 8. Logarithm-integral Function, li(x)

Definition:

Formulas 8.240.1 & 8.240.2

x
⎡1−ε dt x
dt ⎤
li(x ) = ∫
dt
, x <1 li( x ) = lim ⎢ ∫ + ∫ ⎥ = Ei(ln x ), x > 1 (Cauchy Principal Value PV)
ln t ε →0
0 ⎣ 0 ln t 1+ε ln t ⎦

Properties:

li(0 ) = 0 li(+ ∞ ) = +∞ li(1) not defined

Relation to Exponential-Integral Function:

li( x ) = Ei(ln x ), x < 1

( )
li x a = Ei(a ln x ), x < 1, a ≠ 0

Relation to Incomplete Gamma Function:

⎛ 1⎞
− li( x ) = Γ⎜ 0, ln ⎟ = Γ(0,− ln x ) [Formula 8.359.2]
⎝ x⎠
 15

Indefinite Integrals:

∫ ln s = li(s )
ds

References
• Abramowitz & Stegun, Ch. 5.
• Moll, http://www.math.tulane.edu/~vhm/Table.html
• Table of Integrals, Series, and Products, Use of the Tables (“Exponential and related functions”), Sects.
8.24 & Sects. 3.04-3.05 (“Improper Integrals”; “Principal Values”)
• Weisstein, Eric W. "Logarithmic Integral." From MathWorld--A Wolfram Web Resource.
http://mathworld.wolfram.com/LogarithmicIntegral.html
 16

EULER’S CONSTANT
(C or γ )
− e − x ln(x)

Figure 9. Euler’s Constant

Definition:

⎡ n−1 1 ⎤ ⎛ 1 1 1 ⎞
γ = lim ⎢∑ − ln n⎥ = lim⎜1 + + + K + − ln n ⎟ [Formula 8.367.1]
n →∞
⎣ k =1 k ⎦ n →∞
⎝ 2 3 n ⎠

Integral Representations:
∞
⎛1⎞
1
γ = − ∫ exp(− x) ln( x)dx = − ∫ ln ln⎜ ⎟dt [Many other integral representations]
0 0 ⎝t⎠

Properties:

C or γ = 0.577 215 664 ...

Relation to Complete Gamma Function and Psi (Digamma) Function:

Γ ′(1) = Ψ (1) = −γ

References
• Abramowitz & Stegun, Ch. 23.
• Bers, Calculus, pp. 512-3
• Moll, http://www.math.tulane.edu/~vhm/Table.html
• Table of Integrals, Series, and Products, Use of the Tables, and Sect. 8.367 & Index
• Weisstein, Eric W. "Euler-Mascheroni Constant." From MathWorld--A Wolfram Web Resource.
http://mathworld.wolfram.com/Euler-MascheroniConstant.html
 17

CATALAN’S CONSTANT
(G or K )
arctan( x )
x

Figure 10. Catalan’s Constant

Definition:

G=∑
∞
(− 1)m =
1 1 1 1
− 2 + 2 − 2 +K [Formula 0.234.3]
m =0 (2m + 1) 2 2
1 3 5 7

Integral Representations:

arctan( x )
1
G=∫ dx [Many other integral representations]
0
x

Properties:

G or K = 0.915 965 594 ...

References
• Abramowitz & Stegun, Ch. 23.
• Dwight, Table of Integrals, 1961
• Marichev, Oleg; Sondow, Jonathan; and Weisstein, Eric W. "Catalan's Constant." From MathWorld--
A Wolfram Web Resource. http://mathworld.wolfram.com/CatalansConstant.html
• Spiegel, Ch. 34
• Table of Integrals, Series, and Products, Use of the Tables, Formula 4.531.1, & Index
 18

PARTIAL FRACTIONS
Elementary Identities
a
=
a t
=
t bt 1 t−a
m =
x+t ⎛
t ⎜1 +
x⎞ x(a ± bx ) ax a (a ± bx ) t−a t−a
⎟
⎝ t⎠ t t t 1 1 a
=± m = +
a
=
a x( x ± a ) ax a(x ± a ) x x+a x+a x x+a
x−t ⎛x ⎞
t ⎜ − 1⎟ 1 1 1 1 1 a
⎝t ⎠ =± m = −
x( x ± a ) ax a(x ± a ) x x−a x−a x x−a
x a 1
= 1− = 1 1 1
a+x a+x a = − x a + bx a
1+ t (t − 1) t −1 t = −
x a + bx b b a + bx
x a 1 1 1 1
= −1 = = + x a a − bx
a−x a−x a t (1 − t ) 1− t t = −
−1 a − bx b a − bx b
x 1 1 1
= − a + bx
u (t − u ) ut
b a
x
= 1+
a
=
1 t (t − u ) = +
x−a x−a a x a + bx x a + bx
1− 1 1 1
= −
u (t − u ) a − bx
x a b
t (u − t ) ut = −
t 1 1
= 1− = 1 1 1 x x a − bx a − bx
t +1 t +1 1 = −
1+ 2
t x ( x + 1) x x +1 t
=1−
u
=
1
t +u
2
t +u
2
u
x
= 1+
1
=
1 u
=
1
−
1 1+
x −1 x −1
1−
1 t (t − u ) t −u t t2
x t2 u 1
u
=
1
+
1 = −1 =
t (u − t ) u −t u−t
x 1 1 2 2
u −t u
= −1 = t −1
1− x 1− x 1 t2
−1 x 1 a
x = − t2 u 1
t x 1 a + bx b b( a + bx) =1+ =
= 1− = t −u
2
t −u
2
u
x+t x+t x x 1 a 1−
1+ = + t2
t bx − a b b(bx − a )
t 1 a+ x−t
x a−x x a 1 = −
a
= 1−
a a − bx
=
b( a − bx )
−
b
(a + x ) 2
a+x (a + x )2
a x−a x 1 a
= 1− a 1 = −
x x
a + bx
=
bx (a + x ) 2
a+x (a + x )2
1 1+
x= a x 1 a
x −1 = −
1−
x
a
=
1 (a + bx )2 b(a + bx ) b(a + bx )2
±n a − bx bx x a 1
⎛ a⎞ 1− = −
(a ± x ) ±n ±n
= x ⎜1 + ⎟ a (a − bx ) b(a − bx ) b(a − bx )
2 2
⎝ x⎠
⎛ ⎞ ⎛ ⎞
1 ⎜ 1 ⎟⎟
⎜ 1 ⎜ 1 ⎟⎟
⎜
⎛ x⎞
±n x2 x x
= a ± n ⎜1 + ⎟ 1 = +
⎝ a⎠
= = x −a
2 2
2( x − a ) 2( x + a )
ax + b ax ⎜ b ⎟ b⎜ a ⎟
⎜1+ ⎟ ⎜1+ x ⎟
⎝ ax ⎠ ⎝ b ⎠
 19

⎛ ⎞ ⎛ ⎞ x2
=
x
−
x
= ax ⎜ 1 ⎟ = ⎜ ax1 ⎟ a 2 − x2 2( a− x ) 2( a+ x )
±n 1 1 1
(x − a )±n = x ± n ⎛⎜1 − a ⎞⎟ ax−b ⎜ 1− b ⎟ b ⎜ −1 ⎟ x = x + x
⎝ x⎠ ⎝ ax ⎠ ⎝ b ⎠
a 2 − x 2 2 a ( a + x ) 2 a( a − x )
⎛x ⎞
±n
= a ⎜ − 1⎟
±n x
( ) ( )
= a 1 − b 1
( x+ a ) ( x+b ) a −b x+ a a −b x+b
x = x − x
x 2 − a 2 2 a( x − a ) 2 a ( x + a )
= a ( 1 )+ b ( 1 )
⎝a ⎠ x 1 = 1 1
⎛ a ⎞ ( x + a )( x −b ) a+b x+a a+b x−b −
x 2 − a 2 2 a( x − a ) 2 a ( x + a )
a ± bx = bx⎜ ± 1⎟
⎝ bx ⎠ x = a ( 1 )+ b ( 1 ) 1 1 1
( x−a ( x+b ) a +b x−a a+b x+b
) = +
a 2 − x 2 2 a ( a + x ) 2 a( a − x )
⎛ bx ⎞
= a ⎜1 ± ⎟
⎝ a⎠

NOTE: These identities often simplify integral calculations

¾ See Sects. 2.32 & 3.363, below, for examples

dx
¾ Example: GR Formula 2.313.1, I = ∫ a + be mx
.

Change of variable:
s = e mx , ds = me mx dx = msdx
1 ds
I= ∫
m s (a + bs )
1 t t bt
The fraction, , is simplified by above partial fraction, = − ,
s (a + bs ) x(a + bx ) ax a(a + bx )
1 1 ⎛1 b ⎞
with t = 1 , so that = ⎜ − ⎟.
s (a + bs ) a ⎝ s a + bs ⎠
1 ⎛1 b ⎞ 1 ⎛ ds b ⎞
Then, I = ∫ ⎜ −
ma ⎝ s a + bs ⎠
⎟ds = ⎜∫ − ∫
ma ⎝ s a + bs ⎠
ds ⎟.

Now, apply change of variable, t = a + bs, dt = bds, which leads to solution,

I =∫
dx
a + be mx
=
1 ⎛ ds
⎜∫ − ∫ ⎟ =
ma ⎝ s
dt ⎞ 1
t ⎠ ma ma ⎝ t ⎠ ma
[ (
(ln s − ln t ) = 1 ln⎛⎜ s ⎞⎟ = 1 mx − ln a + bemx )]
References
• Bers, Calculus, pp. 421 ff.
• Table of Integrals, Series, and Products, Sect. 2.1 (Rational Functions).
 20

MISCELLANEOUS

Completing The Square

⎛ b ⎞
ax + bx + c = a⎜ x + ⎟ −
2 b 2 − 4ac
2
( ) [The second form is most useful for integrals.
⎝ 2a ⎠ 4a See Sects. 2.32 & 3.323, below, for uses]
⎛
= ⎜⎜ a x +
b ⎞
⎟⎟ −
2
b 2 − 4ac( )
⎝ 2 a⎠ 4a

Finite Binomial Expansions

n
⎛n⎞ n
• (a + b ) = ∑ ⎜⎜ ⎟⎟a n − j b j = ∑
n!
a n− j b j
n

j =0 ⎝ j ⎠ j = 0 j!(n − j )!

n(n − 1) n − 2 2 n(n − 1)(n − 2) n −3 3

= a n + na n −1b + a b + a b + K + bn
2! 3!
n
⎛n⎞ n
• (a − bx ) = ∑ ⎜⎜ ⎟⎟(− 1) a n − j (bx ) = ∑ (− 1) a n − j (bx )
n j j j n! j

j =0 ⎝ j ⎠ j =0 j!(n − j )!
n(n − 1) n − 2 n(n − 1)(n − 2) n −3
= a n − na n −1bx + a (bx ) − a (bx ) + K + (bx ) ,
2 3 n

2! 3!
⎛n⎞
where ⎜⎜ ⎟⎟ is the binomial coefficient, defined as,
j ⎝ ⎠
⎛n⎞ n! n(n − 1)(n − 2 )L(n − j + 1)
⎜⎜ ⎟⎟ = = ,
⎝ j ⎠ j!(n − j )! 1 ⋅ 2 ⋅ 3L j
for factorial n!= n(n − 1)(n − 2)L3 ⋅ 2 ⋅1 = 1 ⋅ 2 ⋅ 3 ⋅ L(n − 2)(n − 1)n and, by definition, 0!= 1!= 1

NOTE: Stirling asymptotic factorial Formula: n!~ n n e − n 2πn

Double Factorial
⎧2 ⋅ 4 ⋅ 6L (n − 2)n for n > 0, even (2n + 1)!!= 1 ⋅ 3 ⋅ 5L(2n − 1)(2n + 1)
⎪ (2n − 1)!!= 1 ⋅ 3 ⋅ 5L(2n − 3)(2n − 1)
n!!= ⎨1 ⋅ 3 ⋅ 5L (n − 2)n for n > 0, odd
⎪1
⎩ for n = 0,−1 (2n )!!= 2 ⋅ 4 ⋅ 6L(2n − 2)(2n)
Note that, by definition, 0!!= −1!!= 1

• See Wolfram Website, http://mathworld.wolfram.com/DoubleFactorial.html

• O’Brien, http://www.docstoc.com/docs/5606124/Double-Factorials-Selected-Proofs-and-Notes

Natural Number N: meaning differs across fields and textbooks to mean all positive integers
(1,2,3,...) either with or without zero included. Here, zero is included in definition, so N = 0,1,K .
 21

Differentiation Under The Integral Sign

dϕ (a ) dψ (a ) ∂f ( x, a )
ϕ (a) ϕ (a)
f ( x, a )dx = f (ϕ (a), a ) − f (ψ (a ), a )
d
1) ∫
da ψ ( a ) da da
+ ∫
ψ (a)
∂ a
dx

dϕ (a ) ∂f ( x, a )
ϕ (a) ϕ (a)
2)
d
∫ f ( x, a)dx = f (ϕ (a), a ) + ∫ dx [c constant ]
da c
da c
∂a
dϕ ( x )
ϕ ( x)
3)
d
∫ f (t , a)dt = f (ϕ ( x), a ) [c constant ]
dx c
dx
dψ (a ) ∂f ( x, a )
c c
4)
d
∫ f ( x, a)dx = − f (ψ (a ), a ) + ∫ dx [c constant ]
da ψ ( a ) da ψ (a)
∂a
dψ ( x )
c
5)
d
∫ f (t , a )dt = − f (ψ ( x), a ) [c constant ]
dx ψ ( x ) dx
∂f ( x, t )
b b
6)
d
∫ f ( x, t )dt = ∫ dt [a, b constant ]
dx a a
∂x
v( x)

f (t )dt = f (v( x) ) − f (u ( x) )
d dv( x) du ( x)
7) ∫
dx u ( x ) dx dx
dϕ ( x )
ϕ ( x)
8)
d
∫ f (t )dt = f (ϕ ( x) ) [a constant ]
dx a
dx
dψ ( x )
a
9)
d
∫ f (t )dt = − f (ψ ( x) ) [a constant ]
dx ψ ( x ) dx

NOTEs:
• Rule # 1 traditionally is called “Leibnitz’s Rule for Differentiating Integrals” (Form. 0.41).
Rules 2-9 are special cases for one and two variable functions with or without constant lower/upper
dΓ( x )
limits of integration. Example—Rule # 6 provides with f ( x, t ) = t x −1e − t , a = 0, b = ∞ limits;
dx
¾ Some integrals cannot be differentiated using these rules, such as—
∞
γ (a, x ) & Γ(a, x ). For example, no rule seems to apply to
d d
∫ exp(−t
2
)dt or
0
da da
dγ (a, x ) d
x ϕ (x )

∫ f (t , a )dt . These forms must be transformed to equivalent

d
da
= ∫
da 0
t a −1 exp(−t )dt =
da c
integrals for which Leibnitz’s rules do apply. See Sects. 3.331 & 8.356 for derivations.

References

• Bers, Calculus, Vol. 2, pp. 808 ff. & Boas, Ch. 4, pp. 233-236
 22

SYMBOL MEANING
int (x ) The integer part of the real number x
(a )n Γ(a + n ) Γ(1 − a )
= a(a + 1)(a + 2)L(a + n − 3)(a + n − 2)(a + n − 1) = = (− 1)
n

Γ(a ) Γ(1 − a − n )
(Pochhammer symbol)
n
= um + um +1 + K + un .
∑u k n
If n < m, we define ∑ uk = 0
k =m

k =m
n = f (m) f (m + 1)L f (n).
∏ f (k ) n
k =m
If n < m, we define ∏ f(k) = 1
k =m

Γ(1 − a )
¾ Suggested new Pochhammer relation: (a )n = (− 1)n is cited in Integrals and Series,
Γ(1 − a − n )
Vol. 1, Elementary Functions, A. P. Prudnikov, et al., 1986, page 772.
¾ Proof :
x =1− a
Denominator Γ(1 − a − n ) = Γ(x − n ) =
(− 1)n Γ(x ) (by Sect. 8.331, below ) = (− 1)n Γ(1 − a ) ,
(1 − x )n (a )n
so that,
(− 1) Γ(1 − a )
n
=
(− 1) Γ(1 − a )
n
= (a )n .
Γ(1 − a − n ) (− 1)n Γ(1 − a )
(a )n
♦ Pochhammer Symbol derivations
O’Brien, http://www.docstoc.com/docs/5473276/Pochhammer-Symbol-
Selected-Proofs
 23

INTRODUCTION
24
 25

Finite Arithmetic Progression

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
Change Formula 0.111 to read: Last form is suggested addition to Formula 0.111.
Last term in sum is
n −1
l−a
∑ (a + kr ) = n [2a + (n − 1)r ] = n (a + l ) l = a + (n − 1)r ⇒ n = + 1.
k =0 2 2 r
l−a
+ 1 into (a + l ) provides the
n
= (a + l )(l − a + r )
1 Substituting
2r r 2
[l = a + (n − 1)r is the last term] proposed alternative solution for the finite sum of
an arithmetic progression, independent of
knowledge of n.
This new form (not seen in common
schoolbooks) is useful when the number of terms
n in a finite progression in unknown, such as,
r =3 (3336 )(3333)
3 + 6 + 9 + K + 3333 = = 1,853,148,
2(3)
giving n = 1,111 terms.

—Formulas 0.111 & 0.112 & 0.121—

 26

Finite Geometric Progression

FORMULA PROOF OUTLINE

(formula references are to G & R 6e)
Change Formula 0.112 to read: Last form is suggested addition to Formula
0.112.
n
( )
a q n − 1 ql − a Last term in sum is aq n−1 = l . lq = aq n is
∑ aq k −1 =
k =1 q −1
=
q −1 substituted into
[q ≠ 1, l = aq n −1
is the last term ] aq n − a
q −1
⇒
ql − a
q −1
. This new form (derived in

many schoolbooks) is useful when the

number of terms in a finite geometric
progression is unknown.

Change Formula 0.121.1 to read: New comment refers reader to more general
expression for arithmetic progression.
n
n(n + 1)
∑k =
k =1 2
[cf. 0.111]

—Formulas 0.111 & 0.112 & 0.121—

 27

INDEFINITE INTEGRALS OF
ELEMENTARY FUNCTIONS
28
29
 30

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
1
1
∫ x e dx = γ ∫
s γ −1e ± s ds
m ± ax n
ds ⎛ s ⎞n m +1
na Set s = ax , = nax n−1 , x = ⎜ ⎟ , γ =
n
.A
m + 1⎤
dx ⎝a⎠ n
⎡
⎢ s = ax , γ = n ⎥
n
useful transform for elementary integrals of
⎣ ⎦ this form.
e ± ax 1− m⎤
n
1 ⎡ Similar to above
∫ x m dx = na z ∫ s e ds
z −1 ± s
⎢⎣ s = ax , z = n ⎥⎦
n
1
integral, ∫ x m e ± ax dx = γ ∫ s γ −1e ± s ds .
n

na

——Sect. 2.31——
 31

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)

( ) ( ) [ln a > 0]
Ei a x Change of variable:
∫ exp a dx =
x

ln a s = a x = e x ln a , ds = a x ln adx = s ln adx
1 es ( )
Ei a x
ln a ∫ s
I= ds =
ln a
by definition of exponential-integral function, Ei( x) ,
Math. Summary.

( )
NOTE: ∫ exp a dx, ∫ xn ( )
exp a x
n

dx do not seem easily

x
solved;math.com, Mathematica not able to solve. But try
setting
s = a x = e x ln a , ds = snx n −1 ln adx which may be reduced
n n

by integration by parts.
¾ verified math.com
ax (− 1)a
• Change of variable:
∫ x a dx = (ln a )1−a Γ(1 − a,− x ln a ) s = a x = e x ln a , ds = a x ln adx = s ln adx, ln s = ln a x
[ln a > 0] ⎛ ln s ⎞
= x ln a, x = ⎜
a

⎟
a

⎝ ln a ⎠
to give,
ax (− 1)b
s ds (ln a )
∫ x b dx = (ln a )1−b Γ(1 − b,− x ln a )
a
1 ds 1 ds
I= ∫
ln a x s a
= ∫
ln a (ln s ) (ln a )
a
= 1− a ∫
(ln s )a
.
[ln a > 0] o Second change of variable:
ds
t = ln s, dt = , e t = s , to give:
s
1 et
( ) (ln a )1−a ∫ t a
2x 2x I=
∫ x2 dx = (ln 2 )li 2 x
−
dt .
x
ƒ Now use result of 2.325.6,
e ax n z +1 z
(
(− 1) a Γ − z,−ax n )
∫ xm dx =
n
(− 1) az +1 z ∞
1 −t m −1
=
n ∫ t z +1
e dt z =
n
− ax n

with parameters [m = a, a = n = 1, z = a − 1] ,
giving,

I=
(− 1)a Γ(1 − a,−t ) = (− 1)a Γ(1 − a,− x ln a )
(ln a )1−a (ln a )1−a
NOTE: derivable directly from 2.325.6 with parameters
[a = ln a, m = a, n = 1].

—Sect.2.312—
 32

•
ax
2 integral, ∫ b dx =
nd (− 1) Γ(1 − b,− x ln a ) similar
b

x (ln a )1−b
to above, and more general
• 3rd integral substitutes a = b = 2 into
ax (− 1)b Γ(1 − b,− x ln a ) = ln 2Γ(− 1,− x ln 2) =
∫ xb dx =
(ln a )1−b
⎡ e x ln 2 ⎤ ⎡ x 2x ⎤
= ln 2⎢Ei( xln2) − ⎥ = ln 2⎢li 2 − ( ) ⎥
⎣ x ln 2 ⎦ ⎣ x ln 2 ⎦

( )
= ln 2li 2 x −
2x
x
by Γ(− 1,− x ) , in Sect. 8.352, below.
¾ Verified math.com
1 ⎛1 ⎞ • First Integral
∫a dx = − Γ⎜ ,− x n ln a ⎟
xn
• Change of variable:
n(− ln a ) ⎝n ⎠
1
n
s = a x = e x ln a , ds = nx n −1 ln a x dx = nx n −1 s ln adx,
n n n

[n ≠ 0, ln a < 0] 1 1
1−
⎛ ln s ⎞ n n −1 ⎛ ln s ⎞ n
ln s = x ln a, x = ⎜
n
⎟ ,x = ⎜ ⎟ , to give,
dx 1 ⎛1 ⎞ ⎝ ln a ⎠ ⎝ ln a ⎠
∫a =− Γ⎜ , x n ln a ⎟
⎝n ⎠
n(ln a )
xn 1
1 s 1 1
∫ ∫
n
I= ds = ds =
[n ≠ 0, ln a > 0] n ln a sx n −1
n ln a x n −1
1 1
1 ∫
ds.
∫x a x dx = n(ln a ) n (ln s ) n
n 1
m 1−

1 ⎛ m +1 ⎞ • Second Change of variable

− m +1
Γ⎜ ,− x n ln a ⎟ ds
n(ln a ) n
⎝ n ⎠ t = ln s, dt = , e t = s, to give,
s
[n ≠ 0, ln a > 0] 1 s 1 et
1 ∫ 1 ∫
I= dt = dt.
n(ln a ) n t n n(ln a ) n t n
1 1
1− 1−

o Evaluate I with Formula 2.325.6,

e ax n
( (
− 1) a z Γ − z ,− ax n
z +1
)
∫ x m dx = n
(− 1) az +1 z ∞
1 −t m −1
=
n ∫ t z +1
e dt z=
n
− ax n

⎡ 1 1⎤
with parameters ⎢a = n = 1, m = 1 − , z = − ⎥ , to
⎣ n n⎦
give:
(− 1) n ⎛ 1
1
1−
⎞ 1 ⎛1 ⎞
I= Γ⎜ ,− x n ln a ⎟ = − Γ⎜ ,− x n ln a ⎟
n(ln a ) n ⎝ n ⎠ n(− ln a ) n ⎝ n ⎠
1 1

NOTE: As a check on calculations, if n = 1, then

—Sect. 2.312—
 33

ax
Γ(1,− x ln a ) =
1 1 x ln a
I =− e = = ∫ a x dx.
(− ln a ) ln a ln a
by Γ(1,− x ) = e , Sect. 8.352, below.
x

• Second and 3rd integrals similar to first

NOTE: integrals derivable from formula, Sect. 2.32,
below,
(− 1)1−γ Γ γ ,−ax n ( )
∫ =
m ax n
x e dx
na γ
(− 1)1−γ ∞ t γ −1e −t dt m +1
= γ
na − ax n ∫ γ=
n
,

(GR 7 errata submission).

NOTE: See Sect. 3.311, below, for definite integrals.

¾ verified math.com
• Re-express and change of variable:
∫a
x2
dx =
1 π
(
erfi ln a x ) [ ln a x ]
2

2 ln a I = ∫ a dx = ∫ e
x2 x 2 ln a
dx = ∫ e dx;
[ln a > 0]
s = x ln a , ds = ln a dx ,
erf (ix )
where erfi( x) = 1
to give, I = ∫
2
i e s ds .
ln a

( )
dx 1 π o Since the indefinite integral for imaginary error
∫a x2
=
2 ln a
erf ln a x function, erfi, is,

[ln a > 0] erfi(t ) =

2
∫ e dt ,
t2

π
then,

I=
1
ln a
∫e
s2
ds =
1 ⎛ π⎞
⎜ ⎟erfi x ln a
ln a ⎜⎝ 2 ⎟⎠
( )
=
1 π
2 ln a
erfi x ln a .( )
NOTE: See Math. Summary (ERROR FUNCTION)
NOTE: See Formula 2.33.2, where by change of variable
(s = a x, ds = a dx , )
∫ e dx = ∫ e
ax 2
( )2 dx s ==a x 1 π erfi a x .
ax

2 a
( )
• nd
2 integral similar to first using Formula 2.33.16
1
∫ ds .
−s 2
for error function, erf, calculation, e
ln a
• verified mathematica

—Sect. 2.312—
 34

NOTE: All solutions in this section can be verified by differentiating under the integral sign. Three
examples follow.
Ei(a x )
• First example— To verify the solution, ∫ exp(a x )dx = , we transform to a one-variable
ln a
definite integral and differentiate the integral. One approach: from Math. Summary,
( )
Ei a x ( )
li e a
x
1 ⎛⎜ dt ⎞⎟
e ax

( )
φ (x)

ln a ⎜ ∫0 ln t ⎟ ∫c f (t )dt. Thus, using Rule # 8

= = x < 0 which is of the form,
ln a ln a
⎝ ⎠
(with Φ(x) = e ) for differentiating integrals [Math. Summary], and the chain rule,
ax

d ⎡ Ei(a )⎤ 1 d 1 [e
1 ⎡ 1 d (e )⎤ = ]
x
ea
1 ⎡e d a ⎤ = ( )
x
dt
⎥ = ln a dx ∫ t = ln a ⎢
ln (e ) dx
ax ax ax
a x ln a = exp a x .
ln a(a ) ⎣ dx ⎦ ln a (a )
x
⎢ ⎥ x
⎢ ⎥ x x
⎣ ⎦
a
dx ln a 0
ln ⎣ ⎦
NOTE: The Principal Value solution for x > 0 is left as an exercise. Hint—use Rule # 9, where 2nd term
⎛ ∞
⎞
in PV definition ⎜ of limiting form, ∫ f (t )dt ⎟⎟ is assumed to differentiate to 0, so that we solve only the
⎜
⎝ 0 ⎠
first term,
d Ei a x d ⎧⎪( ) ⎡ 1 ε e −t ⎤ ⎫⎪
dt ⎥ ⎬ (x > 0).
dx ⎪⎩ ε →0 ⎢⎣ ln a −∫a x t
= ⎨− lim ⎢
dx ln a ⎥⎦ ⎪⎭

• 2nd ex.—Verify solution, ∫

ax
dx =
(− 1) Γ(1 − a,− x ln a ) .
a

x a
(ln a )1− a
∞
Restating Γ(1 − a,− x ln a ) = ∫t
−a
e −t dt , from Math. Summary, and using Rule # 9 for differentiating
− x ln a

integrals [Math. Summary],

d ⎡ (− 1) ⎤ (− 1)
∞
⎡ ⎤ a
a a x

⎢ ∫ t e dt ⎥⎦ = (ln a )1− a
− a −t
⎢⎣ − (− x ln a )− a x ln a d
e (− x ln a ) =
⎥⎦ x a
dx ⎣ (ln a )1− a − x ln a
dx

• 3rd ex. —Verify solution, ∫ a x dx =

1 π
2

2 ln a
( )
erfi ln a x [ln a ≠ 0] . From Math Summary

(Incomplete Gamma Functions), restate solution as,

1 π ⎛⎜ 2 ⎞
1 π
( )
ln a x ln a x
⎟= 1
erfi ln a x = ∫ ∫
t2 2
e dt e t dt ,
2 ln a 2 ln a ⎜⎝ π 0 ⎟
⎠ ln a 0
so that, by Rule # 8 for differentiating integrals [Math. Summary], and laws of exponents,
1 ⎡ ( ln a x )2 d ln a x ⎤
ln a x
d 1
∫ = ⎥=e = ax
t2 x 2 ln a 2
e dt ⎢e
dx ln a 0 ln a ⎣ dx ⎦

————
Note: See Section 8.356, below, for derivatives of algebraic-exponential function integrals
¾ Exercise: Verify other solutions in this section.

—Sect. 2.312—
35
 36

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
•
∫e
± μx
ln p xdx = Integration by parts on each
integral with, u = ln x p , dv = e ± μx dx.
1⎡ ln p −1 xe ± μx ⎤
⎢ ± ln p
xe ± μx
m p ∫ x dx⎥⎥ [μ ≠ 0] NOTE: see def. integrals below in
μ ⎢⎣ ⎦ Sects. 4.331, 4.335.
¾ Mathematica could not do calculation—
“Mathematica could not find a formula for
your integral. Most likely this means that no
formula exists.”
• Set p = 1 in above integral
∫e
± μx
ln xdx =
1⎡ ln p-1 xe± μx ⎤
[± ln xe ] [μ ≠ 0] ∫ e ln xdx = ∫ x dx⎥⎦
± μx ± μx
1
m Ei(± μx )
p
± p
xe m p
μ ⎢⎣
± μx ln
μ
and exponential integral, Ei(± x ) , in
Form. 2.33.18, Form. 2.325.7
o Or—integration by parts on each
integral with, u = ln x, dv = e ± μx dx, and
apply definition of exponential integral,
Ei(± x )
¾ verified mathematica.com
Formula 2.325.4 Derived from expansion Formula (2.324.2) in G
e ax − e ax ae ax a 2 e ax a 3 & R 6e, which see below
∫ x4 dx = − 2 − + Ei(ax )
3x 3 6x 6x 6

•
∫x
n
m
e ax dx = Use Change of Variable & Integration
by Parts.
x m +1− n ax n m + 1 − n m − n ax n
1) Let I = ∫ x m e ax dx;
n

na ∫
e − x e dx
na
ds m +1
set s = ax n , = nax n−1 , γ = , to get,
dx n
1
γ ∫
I= s γ −1e s ds.
na
2) Do integration by parts on ∫ s γ −1e s ds and
re-express in terms of x. Let
u = sγ −1 , du = (γ −1)sγ −2 ds; dv = es ds, v = es ;
uv − vdu = sγ −1es − (γ −1) sγ −2es ds
∫ ∫
Then,
1
na
[
I = γ sγ −1es − (γ −1) sγ −2es ds ∫ ]
(ax ) n γ −1
m +1 − n axn ( ) (nax )e
γ −2 n−1 axn
dx
∫
axn
= γ
e − γ
na n na
Sect. 2.32
 37

x m+1−n ax n m + 1 − n m−n ax n
=
na
e −
na
x e dx ∫
¾ If n = 1, Form. 2.321.1 follows as special
case.
∫x
Formula 2.33.4 m − ax n
Same as above for e dx and signs
∫ x e dx =
n
m ± ax
assigned.
m +1− n NOTE: algebraic-exponential formulas of
x m + 1 − n m − n ± ax n
na ∫ ( )
n
± e ± ax m x e dx the form x ± m exp ± ax n generalize many
na
long-standing formulas with n = 1 and solve
many elementary indefinite and definite
integrals.
(− 1) Γ(γ ,− ax n ) • See Sect. 2.31 for derivation; then apply
1−γ

∫x
m ax n
e dx = definition of incomplete gamma function.
na γ
• Γ(•,•) is incomplete gamma function,
(− 1)1−γ ∞
m +1
Formula 8.350.2, given as Γ(α , x )
∫t
γ −1 −t
= γ
e dt γ =
na n
− ax n ƒ Formula omitted in GR7; sent in as
erratum
ƒ Verify derived integral by differentiating
the integral using Rule # 5 in Math.
Summary with ψ ( x ) = −ax n :
(− 1)1−γ d
∞
f (t , γ )dt
na γ dx ψ ∫( x )
(− 1) ) d (− ax )⎤ = x m e ax
1−γ
⎡
( )
n
e (
γ −1 − − ax n n
= γ ⎢− − ax
n
⎥
na ⎣ dx ⎦

¾ verified mathematica.com
Formula 2.33.5 • Improved Formula 2.33.5 for integer γ .
∫ x e dx =
m ax n
• Derived from Form. 8.352 .4 ,
n −1
xm
e ⎡ γ −1
ax n
⎢ ∑ (− 1)
k (γ − 1)! x n
γ − k −1
⎤
⎥
( ) Γ(n, x ) = (n − 1)!e − x ∑
m = 0 m!
[n = 1,2,...]

n ⎢⎣ k =0 (γ − k − 1)! a k +1 ⎥⎦ • Restate integral as incomplete gamma

function, as previously disclosed (omitted in
⎡ m +1 ⎤
⎢γ = n = 1,2,..., n ≠ 0, a ≠ 0⎥
GR 7):
(−1) Γ(γ ,−axn )
1−γ
⎣ ⎦ m +1
∫x
m axn
e dx = ,γ =
naγ n
• Plug into Form. 8.352.4:
n −1
xm
Γ (n , x ) = (n − 1)! e − x ∑ [ n = 1, 2,...]
m =0 m!
m +1
where substitute d is n = = γ , x = − ax n
n
to give:

Sect. 2.32
 38

γ −1

(−1) (γ −1)!e
1−γ n
ax ∑(− ax ) n k

∫x
m axn
e dx = γ
k =0

na k!

• Now write summation in reverse order and

simplify, noting that its first term is positive,
(− 1)1−γ (− 1)γ −1 = (− 1)0 :

⎢ −
( )
⎡ axn γ −1 axn γ −2 ⎤
⎥
( )
(γ −1)! axn ⎢ (γ −1)! (γ − 2)! ⎥
∫ x e dx= naγ e ⎢ axn γ −3
m axn

⎢+ ( )
−K− (−1)
⎥
1−γ ⎥

⎢⎣ (γ − 3)! ⎥⎦

n ⎢
( ) −
( )
⎡ xn γ −1 (γ −1) xn γ −2 (γ −1)(γ − 2) xn γ −3 ⎤
+ ⎥
( )
eax ⎢ a a 2
a 3
⎥
=
n ⎢
⎢−K−
(γ −1)!(−1) 1−γ ⎥
⎥
⎣ aγ ⎦

e ax ⎡
n
γ −1
(γ − 1)! (x )
n γ − k −1 ⎤
=
n ⎣⎢
⎢ ∑ (− 1) (γ − k − 1)!
k

a k +1
⎥
k =0 ⎦⎥

by inspection, noting relation,

(γ − 1)! k=> 0(γ − 1)(γ − 2 )(γ − 3)K (γ − k ),
(γ − k − 1)!
where the vaule of 1 is understood for k = 0.

NOTE: new version translates more readily

into calculations for γ = 1, 2, 3, 4, given in
Forms. 2.33.6 – 2.33.9.
NOTE: Formula is useful for writing the log-
expansions in Section 2.7 which can be
transformed from log. function to exp.
n
function of the form x m e ax . Three (3)
examples from Sect. 2.7:
1) Form. 2.711, transformed to
exponential form,
∫ ln xdx = ∫ s e ds by change of
m m s

variable
⎛ dx s ⎞
⎜ s = ln x, ds = , e = x ⎟, where
⎝ x ⎠
m is integer
• Applying Form. 2.33.5,

Sect. 2.32
 39

(γ −1)! (xn ) ⎤
γ −k−1
eax
n
⎡ γ −1
⎢ ∑ (−1) k
⎥
n ⎢⎣ k =0 (γ −k −1)! ak+1 ⎥⎦
with
s = x = ln x, n = 1, n → m, γ = m + 1, a = 1,
to give,
m
(−1)k m! (s)m−k .
s =ln x

∫ lnm
xdx = ∫ s m s
e ds = e s
∑
k =0 (m − k )!
m
m!
= x∑ (−1)
k
lnm−k x,
k =0 (m − k )!
which differs from Form. 2.711 only in
expression of factorial term.
2) Form. 2.722,
m ( n +1)s
∫ x ln xdx = ∫ s e ds , by
n m

change of variable
⎛ dx s n+1 (n+1)s ⎞
⎜ s = ln x, ds = , e = x, x = e ⎟
⎝ x ⎠
• Applying
e ax ⎡ γ −1
n

⎢ ∑ (− 1)
k (γ − 1)! x n
γ − k −1
⎤
⎥
( )
n ⎣⎢ k = 0 (γ − k − 1)! a k +1 ⎦⎥

with
s = x = ln x, n = 1, n → m,γ = m + 1, a = n + 1,
to give,
s =ln x

∫ x ln xdx =
n m
∫s
m
e (n+1)s ds =

(− 1) m! (s ) k +1
m m−k
( n+1)s
∑
k
=e
k =0 (m − k )! (n + 1)
m!m
ln m−k x
= x n+1 ∑ (− 1)
k
,
k =0 (m − k )! (n + 1)k +1
which differs from Form. 2.722 only in
expression of factorial term.
3) Formula 2.723.1,
s = ln x
( n +1)s
∫x ln xdx = ∫ se
n
ds by
change of variable,
⎛ dx s n +1 ( n +1)s ⎞
⎜ s = ln x, ds = , e = x, x = e ⎟
⎝ x ⎠

• Applying
eax ⎡ γ −1
n

⎢ ∑(− 1)
k (γ − 1)! xn
γ − k −1
⎤
⎥
( )
n ⎢⎣ k =0 (γ − k − 1)! ak +1 ⎥⎦
Sect. 2.32
 40

with
s = x = ln x, n =1, n → m =1,γ = 2, a = n +1,
to give,
s =ln x
( n +1)s
∫ x ln xdx = ∫ se
n
ds
⎡ s 1 ⎤
= e (n +1)s ⎢ − 2⎥
⎣ n + 1 (n + 1) ⎦
⎡ ln x 1 ⎤
= x n +1 ⎢ − 2⎥
.
⎣ n + 1 (n + 1) ⎦
NOTE: Form. 2.321.2, ∫ x e dx, is a
n ax

∫x
m ax n
special case of e dx, with
m = n, n = 1, γ = n + 1 .
Formula 2.33.6 Substitute values of integer γ in 2.33.5
n
e ax ⎡ m +1 ⎤
∫ x e dx =
n

⎢γ = n = 1⎥
m ax

na ⎣ ⎦

Formula 2.33.7
n
e ax ⎛ xn 1 ⎞ ⎡ m +1 ⎤
∫x
ax n
m
e dx = ⎜⎜ − 2 ⎟⎟ ⎢ γ = n = 2⎥
n ⎝ a a ⎠ ⎣ ⎦

Sect. 2.32
 41

Formula 2.33.8
n
e ax ⎛ x 2n 2 x n 2 ⎞
∫x e
ax n
m
dx = ⎜⎜ − 2 + 3 ⎟⎟
n ⎝ a a a ⎠
⎡ m +1 ⎤
⎢γ = n = 3⎥
⎣ ⎦
Formula 2.33.9
n
e ax ⎛ x 3n 3x 2 n 6 x n 6 ⎞
∫ x e dx =
n
m ax
⎜⎜ − 2 + 3 − 4 ⎟⎟
n ⎝ a a a a ⎠
⎡ m +1 ⎤
⎢γ = n = 4⎥
⎣ ⎦
Formula 2.33.2 • Use change of variable, t = a x & indefinite integral
1 π
∫ e dx = 2 a erfi a x
ax 2
[ ] definition of erfi(t ) : erfi (t ) =
2
∫ e dt.
t2

π
1
∫x
m ax n
¾ Erfi derived from e dx with γ = and Formula
2
⎛1 ⎞
8.359.4, γ ⎜ , x 2 ⎟ with x 2 → − z 2 .
⎝2 ⎠
Formula 2.325.13 Complete the square on exponent, use change of variable,
(ax +2bx +c )dx =
∫e and definition of imaginary error function, erfi, from above.
2

• Thus, completing the square and changing the variable:

1 π ⎛ ac − b 2 ⎞ ⎛ b ⎞ ⎛ b ⎞ 4b 2 − 4ac
2
exp ⎜⎜ ⎟⎟ erfi ⎜ a x + ⎟ ax 2 + 2bx + c = a⎜ x + ⎟ − =
2 a ⎝ a ⎠ ⎝ a⎠ ⎝ a⎠ 4a
[a ≠ 0] ⎛
2
b ⎞ b 2 − ac b ds
⎜⎜ a x + ⎟⎟ − ;s = ax + ; = a;
⎝ a⎠ a a dx
2
⎛ ac − b 2 ⎞ ⎛ b ⎞
I = exp⎜⎜ ⎟⎟ ∫ exp⎜⎜ a x + ⎟⎟ dx =
⎝ a ⎠ ⎝ a⎠
1 ⎛ ac − b 2 ⎞ s 2
exp⎜⎜ ⎟⎟ ∫ e ds
a ⎝ a ⎠
which leads to result.

Mathematica solution at http://integrals.wolfram.com/

shows the awesome speed of computer processing, and the
frequent need for post-processing simplification:

CHECK: If b = c = 0 in ∫ e (ax + 2 bx +c ) dx , the integral

2
•
Sect. 2.32
 42

is identical to ∫ e ax dx =
2 1 π
2 a
[ ]
erfi a x of Form.
2.33.2.
• NOTE: see Form. 2.33.1, GR 7, for a similar negative
exponential integral; uses same approach
involving erf ( x ) =
2
π ∫
−t 2
e dt.

• NOTE: Sect. 3.323, below, has many similar definite

integrals
• Verify solution by differentiating the integral
Formula 2.33.10 • See Sect. 2.31 for derivation; then apply definition and
Γ γ , βx n ( ) properties of incomplete gamma function
∫x
− βx n
m
e dx = − • Γ (• ,• ) is incomplete gamma function,
nβ γ
∞ Formula 8.350 2, p. 890 , given as Γ(α , x )
1 m +1
∫t
γ −1 −t
=− γ
e dt γ = • NOTE: this elementary integral solves and creates
nβ βx n
n many integrals in Sects. 3.3-3.4, below.
• Exercises:
o Verify solution by differentiating under the integral
sign
o Show that the solution,
∫ x e dx = −Γ(a, x ), reproduces the definitions in
a −1 − x

Math. Summary of Γ(a ), Γ(a, x ), γ (a, x ); e.g., show

that,
∞ x =∞

∫x e dx = −Γ(a, x )
a −1 − x
= Γ(a ) by indefinite integral
x =0
0
solution and properties of gamma functions
• verified mathematica.com
Formula 2.33.11 • Improved Formula 2.33.11 for integer γ .
Similar in derivation to above integer-expansion,
∫x
m
(
exp − β x n dx = ) e ax ⎡ γ −1
n

( ) (γ − 1)! x n
γ − k −1
⎤ ( )
∫
m ax n
∑
k
x e dx = ⎢ − 1 ⎥
( )
exp − βx n ⎡ γ −1 (γ − 1)! x n
γ − k −1
⎤ ( ) n ⎣⎢ k =0 (γ − k − 1)! a k +1 ⎦⎥
− ⎢∑ ⎥
⎣⎢ k =0 (γ − k − 1)! β
n k +1
⎡ m +1 ⎤
⎦⎥ ⎢⎣γ = n = 1,2,...⎥⎦
⎡ m +1 ⎤
⎢⎣γ = n = 1,2,..., n ≠ 0, β ≠ 0⎥⎦ • Based on incomplete gamma (special case) expansion,
Form. 8.352.4,
n −1
xm
Γ(n, x ) = (n − 1)! e − x ∑ m!
m=0
[n = 1,2,...]

m +1
where substituted is n = = γ , x = βx n
n
• (a) State integral as incomplete gamma function, Form
2.33.10, (b) translate into incomplete gamma (special
case), Formula 8.352 4, noting that all terms in
summation expression are positive:

Sect. 2.32
 43

∫x
m − βx n
e dx =

−
(
Γ γ , βx n
= −
)
(γ − 1)!e −βx
⎡ γ −1 βx n k ⎤
⎢∑ ⎥
n
( )
nβ γ nβ γ ⎢⎣ k =0 k! ⎥⎦
• Writing summation in reverse order, and simplifying
∫ x e dx =
m − βx n

⎢
( ) +
( )
⎡ β γ −1 x n γ −1 β γ −2 x n γ −2 ⎤
⎥
(γ − 1)!e ⎢ (γ − 1)! (γ − 2)! ⎥
− βx n
−
nβ γ ⎢
⎢+ β ( )
γ −3
xn
γ −3

+K+
1 ⎥
⎥
⎢⎣ (γ − 3)! 0! ⎥⎦

n ⎢
( )
⎡ x n γ −1 (γ − 1) x n γ −2
+
( ) ⎤
⎥
e − βx ⎢ β β2 ⎥
=−
n ⎢
⎢+ (γ − 1)(γ − 2) x
n γ −3

+
( )
K +
(γ − 1)!⎥⎥
⎣⎢ β3 β γ ⎦⎥

=
− e − βx
n

⎢∑
( )
⎡ γ −1 (γ − 1)! x n γ − k −1 ⎤
⎥ by inspection, noting
⎢⎣ k =0 (γ − k − 1)! β
k +1
n ⎥⎦
relation,
(γ − 1)! k=>0(γ − 1)(γ − 2)(γ − 3)K (γ − k ),
(γ − k − 1)!
where the vaule of 1 is understood for k = 0.
NOTE: new version translates more readily into
calculations for γ = 1, 2, 3, 4, Formulas 2.33.12 – 2.33.15.
NOTE: new version is easily derived from above solution,
∫ e dx, by setting a = − β , to give,
m ax n
x

e − βx ⎡ γ −1
n

⎢ ∑ (− 1)
(γ − 1)! x n ( )
γ − k −1
⎤
∫x
m − βx n k
e dx = ⎥=
n ⎣⎢ k = 0 (γ − k − 1)! (− β )k +1 ⎦⎥

−
e − βx ⎡ γ −1 (γ − 1)! x n
n

⎢∑
( )
γ − k −1
⎤
k +1 ⎥
n ⎣⎢ k = 0 (γ − k − 1)! (β ) ⎦⎥
where (− 1) cancels and negative sign is taken outside.
k

NOTE: Formula useful for writing the log-expansions in

Section 2.7 which can be transformed from log. function to
n
exponential function of the form x m e − βx ; e.g., see Sect.
2.7, below, for new formula,
(ln x )n dx = s n −(m−1)s , by change of variable
∫ x m ∫ e ds
⎛ dx s 1− m − ( m −1)s ⎞
⎜ s = ln x, ds = , e = x, x = e ⎟.
⎝ x ⎠
Sect. 2.32
 44

(ln x )n dx =
∫ ∫s e (
n − m −1)s
NOTE: m ds , is solved below from
x
equivalent expression,
(ln x )n dx = s n (1− m )s .
∫ xm ∫ e ds
Formula 2.33.12 • Substitute values of integer γ into Formula 2.33.11
n
e − βx ⎡ m +1 ⎤ • Exercises given below, this section.
∫ x e dx = −
n
m − βx

nβ ⎢⎣γ = n = 1⎥⎦
Formula 2.33.13
n
e − βx ⎛ xn 1 ⎞ ⎡ m +1 ⎤
∫x
n
m
e − βx dx = − ⎜⎜ + 2 ⎟⎟ ⎢γ = n = 2⎥
n ⎝ β β ⎠ ⎣ ⎦
Formula 2.33.14
n
e − βx ⎛ x2n 2 xn 2 ⎞
∫ x e dx = −
n
m − βx
⎜⎜ + 2 + 3 ⎟⎟
n ⎝ β β β ⎠
⎡ m +1 ⎤
⎢γ = n = 3⎥
⎣ ⎦
Formula 2.33.15
n
e − βx ⎛ x 3n 3 x 2 n 6 x n 6 ⎞
∫ x e dx = −
n
m − βx
⎜⎜ + 2 + 3 + 4 ⎟⎟
n ⎝ β β β β ⎠
⎡ m +1 ⎤
⎢γ = n = 4⎥
⎣ ⎦
Formula 2.33.16 • Change of variable, t = β x , and definition of
∫e
− βx 2
dx =
1 π
erf [ βx ] x
erf (x ) =
2 β 2
∫ e dt
−t 2

π 0

= erf (t ) and apply limits, noting that

2
π ∫
−t 2
o Set e

erf (0 ) = 0.
• See Form. 2.33.2 for similar case
1 ⎡ e ax ⎤
n n n
e ax e ax Use Integration by Parts.
∫ xm dx = ⎢ −
m − 1 ⎣⎢ x m −1
+ na ∫ x m−n
dx ⎥ n n
Let u = e ax , du = nax n−1e ax dx
⎦⎥
1
dv = x −m dx, v = ∫ x −m dx = −
(m − 1)x m−1
1 ⎡ e ax ⎤
n n
e ax
∴ uv − ∫ vdu = −
⎢ m−1 + na ∫ m−n
dx ⎥
m − 1 ⎢⎣ x x ⎥⎦
Check : If n = 1, then Formula 2.324.1 gives,
e ax 1 ⎡ e ax e ax ⎤
I= ∫ xm
dx = ⎢ −
m − 1 ⎣ x m −1
+ a dx ⎥
x m−1 ⎦ ∫
n
Formula 2.325.5 e − ax
Same as above for ∫ m dx and signs assigned.
1 ⎡ e ax e ± ax ⎤
n n n
e ± ax x
∫ xm dx = ⎢ −
m − 1 ⎢⎣ x m−1
± na ∫ x m−n dx⎥⎥
⎦
Formula 2.325.6 • See Sect. 2.31, above, for derivation; then apply
definition of incomplete gamma function
Sect. 2.32
 45

e ax
n
( (
− 1) a z Γ − z ,− ax n
z +1
) • Γ(•,•) is incomplete gamma function,
∫ x m dx = n Formula 8.350 .2, given as Γ(α , x )
• Verify solution by differentiating under the integral
(− 1) a
z +1 z ∞
1 −t m −1
=
n ∫ t z +1
e dt z =
n
sign
− ax n
Formula 2.325.8 • Improved Formula 2.325.8 for integer z.
e ax n • Derived from Formula 8.352 .3, as given in GR 6 for
∫ xm
dx = real x,
(− 1)n⎡ n −1
k! ⎤
e ax ⎡ z −1 ( z − k − 1)! a k ⎤ Γ(− n, x ) = Γ(0, x ) − e − x ∑ (− 1) k +1 ⎥
n k
⎢
− ⎢∑
n! ⎣ x ⎦
⎥ k =0
n ⎣ k =0 z! x n ( z −k ) ⎦ • Restate integral as incomplete gamma function, as
( )
n
Ei ax previously disclosed in Form. 2.325.6:
+ az
n z! e ax
n
(
(− 1)z +1 a z Γ − z,−ax n , z = m − 1 )
∫ xm dx =
n n
⎡ m −1 ⎤
⎢n ≠ 0, a ≠ 0, z = n = 1,2,..., ⎥ • Plug into:
⎢ ⎥ Γ(− n, x ) =
(− 1)n ⎡Γ(0, x ) − e − x n −1 (− 1)k k! ⎤ , substituting
⎣m = 2,3,... ⎦ n! ⎢⎣
∑ k =0 x k +1 ⎥⎦
m-1
n=z= , x = −ax n ,
n
and noting that Γ(0,− x ) = − Ei( x ) to give:
n
e ax
∫ x m dx =
⎧ (
⎡Γ 0,−ax n − ) ⎤⎫
(− 1)z +1 a z ⎪ (− 1)z ⎢ z −1 ⎥⎪
⎨ ⎢e ax n (− 1)k k! ⎥⎬
⎢ ∑
n ⎪ z!
⎩ ⎣ k =0 − ax n ( )k +1 ⎪
⎥
⎦⎭
(− 1) (− 1) (− 1)a z Ei(ax n )
z +1 z
= −
n z!
(− 1)z +1 (− 1)z a z e ax z −1
(− 1)k k!
∑ (− 1) (−1)a
n

( k +1)
n z! k =0
k k +1
xn

=−
a z ax n z −1
n z!
e ∑
k +
k!
n ( k +1) +
az
n z!
Ei ax n ( )
k =0 a x
1

• Write summation term in reverse order and

simplify,

Sect. 2.32
 46

⎡ (z − 1)! (z − 2)! (z − 3)! ⎤

z ⎢ + z −1 n ( z −1) + z − 2 n ( z − 2 ) ⎥
a ax n a x z nz
a x a x
− e ⎢ ⎥
n z! ⎢+ K + 1 ⎥
⎢⎣ ax n ⎥⎦
⎡ ( z − 1)! ( z − 2 )!a ( z − 3)!a 2 ⎤
+ +
e ax ⎢ z! x nz n ( z −1) ⎥
z! x n ( z −2 ) ⎥
n

⎢ z! x
=−
n ⎢ a z −1 ⎥
⎢ + K + ⎥
⎣ z! x n
⎦

⎡ 1 a ⎤
⎢ zx nz + z ( z − 1)x n ( z −1) +
ax n ⎥
e ⎢ ⎥
=−
n ⎢ a2 a z −1 ⎥
⎢ +K+ ⎥
⎣ z (z − 1)( z − 2 )x
n ( z −2 )
z! x n ⎦
e ax ⎡ z −1 ( z − k − 1)! a k ⎤
n

=− ⎢
n ⎣ k =0
∑ z!
⎥
x n ( z −k ) ⎦
by inspection, noting the factorial relation,
(z − k − 1)! = 1
.
z! z ( z − 1)(z − 2)K( z − k )
NOTE: new version translates more readily into
calculations for z = 1,2,3, Formulas 2.325.9 – 2.325.11.
NOTE: Formula 2.324.2 may be restated from above
integral expansion by letting
m = n, n = 1, z = n − 1, to give :
e ax ax ⎡
n−2
(n − k − 2 )! a k ⎤ a n −1
Ei(ax )
∫ x n = −e ⎢⎣∑ k =0
+
(n − 1)! x n − k −1 ⎥⎦ (n − 1)!
which differs only in the start/end summation limits and
statement of factorial terms.
e ax
NOTE: Formula 2.324.2, ∫ n dx, is a special case of
x
ax m
e m −1
∫ x n dx, z = n , with n = 1, m → n, z = m − 1.
Γ(0,− ax n ) Ei(ax n )
n
Formula 2.325.7 e ax
Derived from, ∫
e ax
n
( )
Ei ax n x
dx = −
n
=
n
,
∫ x dx =
n m −1
in Form. 8.359.1, where z = =0.
n
¾ See Mathematical Summary and Sect. 8.212, below,
for definitions of ± Ei(± x ) in terms of incomplete
gamma function, Γ(α , x ).
Formula 2.325.9 Substitute values of integer z in 2.325.6

Sect. 2.32
 47

n n
e ax − e ax aEi(ax n )
∫ x m dx = nx n + n
⎡ m -1 ⎤
⎢ z = n = 1⎥
⎣ ⎦

Formula 2.325.10
n n n
e ax − e ax ae ax a 2 Ei(ax n )
∫ xm dx = −
2nx 2 n 2nx n
+
2n
⎡ m -1 ⎤
⎢ z = n = 2⎥
⎣ ⎦

Formula 2.325.11
n
e ax
∫ x m dx =
n n n
− e ax ae ax a 2 e ax a 3 Ei(ax n )
− − +
3nx 3n 6nx 2 n 6nx n 6n
⎡ m -1 ⎤
⎢ z = n = 3⎥
⎣ ⎦

Formula 2.325.12 Reduction formula 2.325.5 and apply 2.33.2,

[ ] ∫e
2 2
e ax e ax ax 2
dx.
∫ x 2 dx = − x + aπ erfi a x
Formula 2.33.19 • See Sect. 2.31 for derivation; then apply definition of
e − βx
n
(
β z Γ − z, βx n ) incomplete gamma function
∫ x m dx = − n o Set m → − m in ∫x
m − βx n
e dx
βz ∞
e −t m −1 Γ(•,•) is incomplete gamma function,
n β∫x n t z +1
=− z= •
n Formula 8.350.2, p. 890 , given as Γ(α , x )
• Useful formula for solutions in 3.3-3.4, below.
• Verify solution by differentiating under the integral
sign
• verified mathematica.com
Formula 2.33.19 • Improved Formula 2.33.19 for integer z
• Derived from Formula 8.352 .3, as given in GR 6 for
real x,
Γ(− n, x ) =
(− 1)n ⎡Γ(0, x ) − e − x n−1 (− 1)k k! ⎤
n! ⎢⎣
∑
k =0 x k +1 ⎥⎦
• Restate integral as incomplete gamma function, as
previously disclosed in 2.33.17:
e − βx
n
(
− β z Γ − z, βx n )
m −1
∫ xm dx =
n
,z=
n
• Plug into:
Sect. 2.32
 48

(
exp − βx n ) Γ(− n, x ) =
(− 1)n
⎡ n −1
k! ⎤
Γ(0, x ) − e − x ∑ (− 1) k +1 ⎥ ,
∫ xm dx =
k
⎢
n! ⎣ x ⎦
k =0

(
exp − βx n ⎡ z−1) ( )k ( z − k − 1)! β
k
⎤ substituting n = z =
m-1
, x = βx n , and noting that
− ⎢∑ − 1 n( z −k ) ⎥
+ n
n z! x
⎣ k =0 ⎦ Γ(0, x ) = − Ei(− x ) , to give:
(−1)z β ( )
z n

Ei − βx n e − βx
nz! ∫ x m dx =
⎡ m −1 ⎤
⎢ z = n = 1,2,..., m = 2,3,...,⎥ − β z ⎧⎪ (− 1) ⎡ (− 1)k k! ⎤ ⎫⎪
z
( )
z −1

⎨ ⎢ Γ 0, β x n
− e − βx n
∑ ⎥⎬
⎢
⎣n ≠ 0, β ≠ 0
⎥
⎦
n ⎪ z!
⎩ ⎢⎣ k =0 β x
n k +1
( )
⎥⎦ ⎪⎭
− (− 1) β z
z
=
n z!
− Ei − β x n − ( )
(− 1)(− 1)z β z e − βx z −1
(− 1)k k!
∑β
n

k +1
n z! k =0 x n ( k +1)
(− 1)z β z Ei(− βx n ) + (− 1)z β z e − βx z −1
(− 1)k k!
∑
n
=
k =0 β
k +1 n ( k +1)
n z! n z! x
• Write summation term in reverse order and simplify,
noting that its first term in oscillating series is negative,
⎡ (− 1)z −1 ( z − 1)! (− 1)z −2 ( z − 2)!⎤
+
(− 1)z β z e −βxn ⎢⎢ β z x nz β z −1 x n ( z −1) ⎥
⎥
n z! ⎢ 1 ⎥
⎢− K − n ⎥
⎣ βx ⎦
⎡ (− 1)2 z −1 ( z − 1)! (− 1)2 z −2 ( z − 2 )! β ⎤
n
⎢ + ⎥
=
e − βx
⎢ z! x nz z! x n ( z −1) ⎥
n ⎢ β z −1
⎥
⎢− K − (− 1)
z
⎥
⎣ z! x n
⎦
k ( z − k − 1)! β
n
− e − βx ⎡ z −1 k
⎤
= ⎢ ∑ (− 1) n ( z −k ) ⎥
n ⎣ k =0 z! x ⎦
by inspection, noting relation,
(z − k − 1)! = 1
.
z! z ( z − 1)( z − 2 )K ( z − k )
NOTE: new version translates more readily into
calculations for z = 1, 2, 3, Formulas 2.33.20 – 2.33.22.
n
e ax
NOTE: easily derived from above expansion, ∫ m dx, by
x
setting a → − β , and rearranging, which adds a
(− 1)k multiplier to summation symbol.
Formula 2.33.18 See 2.325.7 for similar case
e − βx
n
(
Ei − βx n ) ¾ Verify solution by differentiating under the integral
∫ x dx =
n
sign (convert Ei(•) to definite integral.
Sect. 2.32
 49

Formula 2.33.20 Substitute values for integer z in 2.33.19

e − βx n
e β Ei(− βx
− βx n n
)
∫ x m
dx = −
nx n
−
n
⎡ m −1 ⎤
⎢⎣ z = = 1⎥
n ⎦
Formula 2.33.21
n n n
e − βx e − βx β e − βx
∫ x m dx = − 2nx 2n + 2nx n +
(
β 2 Ei − βx n )
2n
⎡ m −1 ⎤
⎢⎣ z = n = 2⎥⎦
Formula 2.33.22
n n n
e − βx e − βx β e − βx
∫ x m dx = − 3nx 3n + 6nx 2n −
β 3 Ei(− βx n )
n
β 2 e − βx
−
6nx n 6n
⎡ m −1 ⎤
⎢⎣ z = = 3⎥
n ⎦
Formula 2.33.23 Integration by parts (Form. 2.325.5), change of variable on
second term, and definition of erf ( x ) , Form. 8.250.1,
e − βx e − βx
[ ]
2 2

∫ x 2 dx = − x − βπ erf β x
2
x

∫e
−t 2
dt .
π 0

Formula 2.321.2 • Proposed alternative restatement of 2.321.2.

ax ⎡ x ⎤
n n−k In above expansion,
n!
∫ x e dx = e ⎢⎣ ∑ (− 1) (n − k )!
n ax k

a k +1 ⎥⎦ e ax ⎡ γ −1
n

( ) ( ( )
γ − 1)! x n
γ − k −1
⎤
∫
m ax n
∑
k
k =0
x e dx = ⎢ − 1 ⎥
n ⎢⎣ k =0 (γ − k − 1)! a k +1 ⎥⎦
⎡ m +1 ⎤
⎢⎣γ = n = 1,2,...⎥⎦
set n = 1, m → n, γ = n + 1 and substitute.
NOTE: proposed formula is essentially the same as
suggested in online GR 6 errata, and published in GR 7,
http://www.mathtable.com/errata/gr6_errata.pdf,
Errata #47, by Drs. Moll and Boros— except we state
n!
for the equivalent expression,
(n − k )!
⎛n⎞ n!
k!⎜⎜ ⎟⎟ = k! ,
⎝k ⎠ k!(n − k )!
n!
noting that, = n(n − 1)(n − 2 )K (n − k + 1).
(n − k )!
Sect. 2.32
 50

Formula 2.324.2 • Alternative expression for Formula 2.324.2:

Set n = 1, m → n, z = n − 1 in above integral,
e ax
∫ x n dx =
n
e ax − e ax ⎡ z −1 ( z − k − 1)! a k ⎤
n

z Ei ax( )
n

∫ xm dx = ⎢∑ ⎥ + a
n ⎣ k =0 z! x n( z −k ) ⎦ n z!
⎡ n−2 (n − k − 2 )! a k ⎤
− e ax ⎢∑ n − k −1 ⎥
+ ⎡ m −1 ⎤
⎣ k = 0 (n − 1)! x ⎦ ⎢ z = n = 1,2,..., m = 2,3,...⎥
⎣ ⎦
a n−1
Ei(ax ) [a ≠ 0]
(n − 1)!
− βx ⎡
n
n! x n −k ⎤ −βx ⎡
n
n! x n−k ⎤
∫ ⎢∑ ∫ ⎢∑ (n − k )! β k +1 ⎥ is derived from
n − βx n − βx
x e dx = − e k +1 ⎥
• x e dx = − e
⎣ k =0 (n − k )! β ⎦ ⎣ k =0 ⎦
[β ≠ 0] general case,
⎛x 1 ⎞ − e − βx ⎡ γ −1 (γ − 1)! x n
n
( )
γ − k −1
⎤
∫ xe
− βx
dx = − e − βx ⎜⎜ + 2 ⎟⎟ ∫ x m − βx n
e dx = ⎢ ∑
n ⎣⎢ k =0 (γ − k − 1)! β k +1 ⎦⎥
⎥
⎝β β ⎠
[β ≠ 0] ⎡ m +1 ⎤
⎢⎣γ = = 1,2,...⎥
⎛x
2
2x 2 ⎞ n ⎦
∫x e
2 − βx
dx = − e − βx ⎜⎜ + 2 + 3 ⎟⎟
⎝ β β β ⎠ by substituting n = 1, m → n, γ = n + 1
[β ≠ 0] • ∫ xe
− βx
dx,∫ x 2 e − βx dx & ∫ x 3 e − βx dx obtained by setting n
−βx ⎛ x 3x2 6x 6 ⎞
3
= 1, 2, 3, respectively, in
∫
3 −βx
x e dx = − e ⎜
⎜ β β 2 + β 3 + β 4 ⎟⎟
+
− βx ⎡ n! x n − k ⎤
n
⎝ ⎠
∫ ⎢∑ (n − k )! β k +1 ⎥.
n − βx
x e dx = − e
[β ≠ 0] ⎣k =0 ⎦
e − βx dx e − βx dx Ei(− βx n )
n

∫ x = Ei(− βx ) From ∫
x
=
n
with n = 1.
[β ≠ 0]
e − βx dx e − βx − βx ⎡ k (n − k − 2 )! β ⎤
n−2 k

∫ xn = ∫ x n dx = −e ⎢⎣∑ (− 1) n − k −1 ⎥
+
k =0 (n − 1)! x ⎦
•
k (n − k − 2 )! β
(− 1)n−1 β z Ei(− βx )
⎡ n−2 k
⎤
e − βx ⎢∑ (− 1)
n
n − k −1 ⎥
+
⎣ k =0 (n − 1)! x ⎦ (n − 1)!
(− 1)n−1 β n−1 Ei(− βx ) [β ≠ 0] is derived from general case,
(n − 1)! k ( z − k −1)! β
n n
e −βx − e −βx ⎡ z−1 k
⎤
∫ xm dx = ⎢∑ (−1) ⎥+
n ⎣k =0 z! x n( z −k ) ⎦

(−1) z β z Ei(− βx )
e − βx dx − e − βx
∫ x 2 = x − βEi(− βx ) ⎡ m −1
n
⎤
⎢ z= = 1,2,..., m = 2,3,...⎥
n z! ⎣ n ⎦
[β ≠ 0] by substituting n = 1, m → n, z = n − 1
e − βx dx e − βx dx e − βx dx
e − βx dx − e − βx β e − βx 2 Ei(− β x )
∫ x3 = 2x2 + 2x + β 2
• ∫ x 2 ,∫ x3 & ∫ x 4 : set n = 2, 3, 4,
[β ≠ 0] e − βx
respectively in ∫ n dx.
e − βx dx x
∫ x4 =
Sect. 2.32
 51

− e − βx β e − βx β 2 e − βx Ei(− βx )
+ − − β3
3x 3
6x 2
6x 6
[β ≠ 0]
∫x x n+1
n
dx
A basic integral equal to (n ≠ −1). Show derivation
n +1
using exponential functions.
One approach:
In integral, I = ∫ x n dx , set x n = e n ln x from definition.
Then, I = ∫ e n ln x dx . Use change of variable,
dx s
s = ln x, ds = , e = x , to give,
x
( n +1)s et
I = ∫e ds leading to I = ∫ dt by a second
n +1
change of variable, t = (n + 1)s , which leads to answer,
et e (n+1)s e ( n +1) ln x x n +1
I= = = =
n +1 n +1 n +1 n +1
by back substitution and laws of exponents.
NOTE: Other approaches are possible once
I = ∫ e (n +1)s ds is obtained.
1 Exercise. Similar to ∫ x n dx .
∫ x n dx
1 1 1
∫ x 2 dx & ∫x 3
dx Exercise. Check against solution above for ∫x n
dx .

∫ x dx
x Computers cannot solve this indefinite integral in closed
form. Try it on mathematica.com. We can write, from
definition,
x x = (e ln x ) = e x ln x . And the derivative is given by the
x

chain rule,
dx x e x ln x d ( x ln x ) x ln x
=d = e = (ln x + 1)e x ln x
dx dx dx
= x (ln x + 1).
x

Substitute and see result. No elementary forms can be

substituted to provide a closed form solution.
See GR, Section 1.211, for series expansion of x x by
looking at a x and setting a → x.
ax n Derived by above formula (with t set to s ) ,
(− 1)
1− γ e
⎛ 1⎞
γ −1

∫x ∫ − 1) Γ(γ ,−ax n )
m ax n
e dx = ⎜ ln ⎟ dt ( 1− γ
na γ
∫x
m ax n
⎝ t⎠ e dx =
naγ
0

m +1
γ= (− 1)1−γ ∞
m +1
∫s
γ −1 − s
n = γ
e ds, γ =
.
na − ax n n
Use change of variable on definite integral,
Sect. 2.32
 52

(− 1)1−γ ∞

∫s
γ −1 − s
γ
e ds :
na − ax n

t = e − s , dt = −e − s ds = −tds
n
t −ax n = e ax , t ∞ = 0,
1
ln t = − s, s = − ln t = ln
t
Then,
(− 1)
1−γ 0
t
∫ (− ln t )
γ −1
I= γ
dt
na n −t
e ax
n

(− 1)
1−γ e ax
⎛ 1⎞
γ −1
=
na γ ∫
0
⎜ ln ⎟
⎝ t⎠
dt

Thus,
(− 1)1−γ Γ(γ ,−ax n ) =
∫ x e dx =
n
m ax

na γ
(− 1)1−γ (− 1)
n
∞ 1−γ e ax γ −1
⎛ 1⎞
∫s ∫
γ −1 − s
e ds = ⎜ ln ⎟ dt ,
na γ − ax n
na γ 0 ⎝ t⎠
m +1
γ= .
n
• Derive similar logarithmic expressions for these 3
algebraic-exponential integrals given earlier:
n n
e ax e − βx
∫ x e dx & ∫ x m dx & x m dx
m − βx n

¾ Verify each solution by diff. under the integral

e ax − 1
( )
2 We show two alternative solutions.
∫ e +1
ax
dx = ln e ax + 1 − x
a
• Separate the terms,
e −1
ax
e ax 1
2 ⎛ 2
⎜
ax
−
ax
⎞
⎟
∫ eax +1 ∫ eax +1dx − ∫ eax +1 dx
dx =
= ln ⎜ e + e 2 ⎟
a ⎝ ⎠ • Change of variable:
s = e , ds = ae ax dx = asdx
ax

Then,
s ds 1 ds 1 ⎡ ds ds ⎤
I =∫ −∫ = ⎢∫ −∫
(s + 1) as (s + 1) as a ⎣ s + 1 s(s + 1)⎥⎦
Use partial fractions for second integral,
1 1 1
= − , to give,
s (s + 1) s s + 1
1 ⎡ ds ⎛1 1 ⎞ ⎤
I= ∫ −∫ ⎜ − ⎟ds
a ⎣ s + 1 ⎝ s s + 1 ⎠ ⎥⎦
⎢

=
1
[ln(s + 1) − ln s + ln(s + 1)] = 1 [2 ln(s + 1) − ln s ]
a a

Sect. 2.32
 53

• At this stage two solutions can be obtained. First

solution is obtained by substituting e ax :
I = [2 ln (s + 1) − ln s ]
1
a
1
[ ( ) ]
= 2 ln e ax + 1 − ln e ax
a
1
[ ( ) ]
= 2 ln e ax + 1 − ax =
( )
2 ln e ax + 1
−x
a a
¾ This is computer solution at mathematica.com
Also, using the log. relation in Math. Summary,
⎛ ⎞
⎜ x ⎟
a ln x − b ln y = a ln⎜ b ⎟, with a = 2, b = 1,
⎜ ya ⎟
⎝ ⎠
the second solution is:
⎛ ⎞
2 ⎜ s + 1⎟
I = [2 ln(s + 1) − ln s ] = ln⎜ 1 ⎟
1
a a ⎜ 2 ⎟
⎝ s ⎠
2 ⎛ 2 − ⎞ 2 ⎛ ⎞
1 1 ax ax
−
= ln⎜⎜ s + s 2 ⎟⎟ = ln⎜⎜ e 2 + e 2 ⎟⎟
a ⎝ ⎠ a ⎝ ⎠
¾ Cited in Prudnikov, et al., Vol. 1, Formula 1.3.1.7, p.
137
e ax + 1 2 ⎛
ax
−
ax
⎞
¾ Exercise: Show ∫ ax dx = ln⎜⎜ e 2 − e 2 ⎟⎟
e −1 a ⎝ ⎠
e ax − m m + n ⎛⎜ m + n
nax
−
m ax
⎞ ¾ One solution solved by separation of terms, change of
∫ eax + n dx =
na
ln ⎜ e + ne m+n ⎟
⎟ variable, s = e ax , partial fraction,
⎝ ⎠
1 1 1
=± m , and relation,
x(x ± a ) ax a( x ± a )
⎛ ⎞
⎜ x ⎟
a ln x − b ln y = a ln⎜ b ⎟ .
⎜ ya ⎟
⎝ ⎠
¾ Computer provides alternative solution,
e ax − m (m + n ) ax( )mx
∫ e ax + n dx = na ln e + n − n .
e ax − 1
¾ Check against integral solution, ∫ ax dx
e +1
with m = n = 1, and by differentiation.
m + n ⎛⎜ m + n ⎞
nax m ax Exercises.
e ax + m −
∫ e −n
ax
dx =
na
ln ⎜ e − ne m+n ⎟
⎟ Similar to above integral,
⎝ ⎠ e ax − m m + n ⎛⎜ m+ n
nax
−
m ax
⎞
[( ] ∫ eax +n m+ n ⎟
e ax + m dx = ln e + ne
1
∫ eax + n dx = na ln e + n e
ax
)
n − m m ax
na ⎜
⎝
⎟
⎠
• Each integral has two alternative solutions. We give
Sect. 2.32
 54

e ax − m 1
[(
n − m m ax
∫ eax −n dx = na ln e − n e
ax
) ] only one solution, based on the log. relations,
⎛ ba ⎞
(
a ln x + b ln y = ln x y = a ln⎜⎜ xy ⎟⎟
a b
)
⎝ ⎠
⎛ ⎞
⎛ xa
⎞ ⎜ x ⎟
a ln x − b ln y = ln⎜⎜ b
⎟⎟ = a ln⎜ b ⎟
⎝y
⎠ ⎜ ya ⎟
⎝ ⎠
¾ The computer gives the second solution.
∫ e dx = e
x x Verify solution to this basic integral by earlier exponential
integral formula,
( )
(− 1)1−γ Γ γ ,−ax n = (− 1)1−γ ∞ t γ −1e −t dt
∫ ∫n
m ax n
x e dx =
na γ na γ −ax
m +1
γ= , with γ = 1, and Γ(n,-x ), Sect. 8.352, below.
n
∫e
−x
dx = −e − x Show by Formula 2.33.11, above,
(
Γ γ , βx n )
1
∞

∫ γ ∫
m − βx n
x e dx = − γ
= − t γ −1e −t dt
nβ nβ β x n
m +1
γ= , with γ = 1, and Γ(n, x ), Sect. 8.352, below.
n
dx (
ln e ax + 1 ) • Use change of variable,
∫ 1 + eax = x −
a s = e ax , ds = ae ax dx = asdx
• Then, by partial fraction expansion and rules of
logarithms,
1 1 ds 1 ⎛ 1 1 ⎞
I= ∫ = ∫⎜ − ⎟ds
a (s + 1) s a ⎝ s s + 1 ⎠

[ln(s ) − ln(s + 1)] = 1 ln⎛⎜⎜ e ax ⎞⎟⎟

ax
1
=
a a ⎝1+ e ⎠
1
[ ( ) (
= ln e ax − ln 1 + e ax = x − )]ln 1 + e ax ( )
a a
e ax • Use change of variable,
∫ x + bdx = e Ei(ax + ab)
− ab
s = x + b, ds = dx, x = s − b
• Then,
ea (s −b ) − ab e
as
I =∫ ds = e ∫ ds = e − ab Ei(as )
s s
= e Ei(ax + ab )
− ab

e ax Ei ax n
n
( )
By Formula 2.325.7,
x
dx =∫ n
, with n = 1.

x +1 − e − ax (ax + a + 1) Exercise. Check answer on mathematica.com

∫ eax dx =
a2
• Uses a special function, the dilogarithm, Li 2 ( x ), defined
x
∫ e x −1 dx = x log 1 − e (
x
−
x2
2
)
+ Li 2 e x( )
Sect. 2.32
 55

∞
zk
as an infinite series, Li 2 (z ) = ∑ , z < 1. The
k =1 k
2

log(1 − x )
indefinite integral is taken as: Li 2 (x ) = − ∫ dx.
x
x
• In I = ∫ dx , use change of variable, s = e x , and
e −1
x

partial fraction expansion to give,

ln s ln s
I =∫ ds − ∫ ds .
s −1 s
The second integral is solved by change of variable,
t = ln s, to give,
t 2 (ln s )
2
ln s x2
∫ s ds = ∫ tdt =
2
=
2
=
2
.
The first integral is more involved. The exercise below
will show that, after integration by parts,
ln s ln s
∫ s − 1ds = ∫ 1 − s ds = ln s ln(1 − s ) + Li2 (s ) ,
so that,
x x2
∫ e x −1 dx = ln s ln (1 − s ) + Li 2 ( s ) −
2
x
which leads to solution when e is substituted,
x
∫ e x −1 dx = x (
log 1 )
− e x
−
x2
2
( )
+ Li 2 e x
NOTE: Do not confuse notation of Li 2 ( x ) with logarithm-
integral function, li( x ).
• Exercise: derive selected common dilogarithm
relations:

Sect. 2.32
 56

ln (1 ± x )dx (ln x )2 − Li ⎛ 1⎞
∫ = −Li 2 (m x ) = 2⎜m x ⎟
x 2 ⎝ ⎠
ln xdx ⎛ x⎞ ⎛ x⎞
∫ = ln x ln⎜ 1 ± ⎟ + Li 2 ⎜ m ⎟
x±a ⎝ a⎠ ⎝ a⎠
ln xdx ln xdx
∫ = ∫ (see above )
a−x x−a
ln (a ± x )dx ⎡ ⎛ x ⎞⎤ ⎛ x⎞
∫ = ln x ⎢ln (a ± x ) − ln⎜ 1 ± ⎟⎥ −Li 2 ⎜ m ⎟
x ⎣ ⎝ a ⎠⎦ ⎝ a⎠
⎛ x ⎞ (ln x )
2
= ln a ln x − Li 2 ⎜ m
⎛ a⎞
⎟= − Li 2 ⎜ m ⎟
⎝ a⎠ 2 ⎝ x⎠
ln ( x − a )dx ⎡ ⎛ x ⎞⎤ ⎛ x ⎞ (ln x ) ⎛a⎞
2

∫ = ln x ⎢ln ( x − a ) − ln ⎜ 1 − ⎟⎥ −Li 2 ⎜ ⎟= − Li 2 ⎜ ⎟
x ⎣ ⎝ a ⎠⎦ ⎝ ⎠
a 2 ⎝x⎠
ln (a ± bx )dx ⎡ ⎛ bx ⎞ ⎤ ⎛ bx ⎞
∫ = ln x ⎢ln (a ± bx ) − ln ⎜ 1 ± ⎟⎥ −Li 2 ⎜ m ⎟
x ⎣ ⎝ a ⎠⎦ ⎝ a⎠
⎛ bx ⎞ (ln bx ) ⎛ a⎞
2
= ln a ln x − Li 2 ⎜ m ⎟= − Li 2 ⎜ m ⎟
⎝ a⎠ 2 ⎝ bx ⎠
ln (bx − a )dx ⎡ ⎛ bx ⎞ ⎤ ⎛ bx ⎞
∫ = ln x ⎢ln (bx − a ) − ln⎜1 − ⎟⎥ −Li 2 ⎜ ⎟
x ⎣ ⎝ a ⎠⎦ ⎝a⎠

=
(ln bx ) 2
− Li 2 ⎜
⎛a⎞
⎟
2 ⎝ bx ⎠
⎡ ⎛ x ⎞⎤ ⎡ ⎛ a ⎞⎤
ln ⎢a⎜⎜1 ± ⎟⎟ ⎥ dx ln ⎢ x⎜⎜ ± 1⎟⎟ ⎥ dx
ln (a ± x )dx ⎣ ⎝ a ⎠⎦ ⎣ ⎝ x ⎠⎦
Ex: ∫ =∫ =∫ ; first
x x x
integral requires two integrations by parts with partial fraction
identity on du term; second and third solutions use log. relation,
log xy = log x + log y, and definition of Li 2 ( x ).
• NOTE: Li 2 ( x ) has useful results important for definite

π2 π2
integrals; e.g., Li 2 (1) = ; Li 2 (− 1) = − . See The
6 12
Wolfram Function Site http://functions.wolfram.com/.
dx dx dx dx
∫ b
&
b ∫
&
a − be mx
• Three exercises. See ∫ a + be mx
derivation in the
a + mx a − mx
e e Mathematical Summary.
• Check solution on mathematica.com
∫x e
3 − βx 2 Derive from Formula 2.33.12, above,
dx
⎡ m +1 ⎤
∫x
m − βx n
e dx ⎢⎣γ = n = 2⎥⎦

∫x e
2 − βx
dx Derive from Formula 2.33.12, above,
⎡ m +1 ⎤
∫x
m − βx n
e dx ⎢⎣γ = n = 3⎥⎦

m +1
• Try other combinations of m, n for γ = = integer.
n

Sect. 2.32
 57

∞ x
e−t
( )
et In this definition, show equivalence of two integrals by
Ei( x ) = − ∫ dt = ∫ dt = li e x , x < 0 stating the indefinite integral forms, substituting the
−x
t −∞
t limits, and applying properties of the exponential-
integral function summarized in Mathematical
ex
Summary. Then show Ei( x ) = li e ( ) = ∫ lndtt
x
by change
0
of variable, s = ln t.
−ε
⎡ e −t
e
∞
⎤ −t Cauchy Principal Value (PV) for definition of exponential-
Ei( x ) = − lim ⎢ ∫ dt + ∫ dt ⎥, x > 0 integral function, x > 0 .
⎣− x t t
ε → +0
ε ⎦ • Show the limit calculation by stating the indefinite
integral forms, substituting the limits, applying
properties of the exponential-integral function in
Mathematical Summary, and evaluating the limit,
ε → +0.
x
dt In this definition, show the relation, li(x ) = Ei(ln x ), x < 1,
li( x ) = ∫ = Ei(ln x ), x < 1 by change of variable,
0
ln t
x
dt
s = ln t on integral, ∫ .
0
ln t
⎡1−ε dt x
dt ⎤ Cauchy Principal Value PV for definition of logarithm-
li( x ) = lim ⎢ ∫ + ∫ ⎥ = Ei(ln x ), integral function, li( x ), x > 1.
⎣ 0 ln t 1+ε ln t ⎦
ε →0
• Perform same operations as for above exponential-
x >1 integral function, Ei( x )
xe x ex • Change of variable,
∫ (1 + x )2 1 + x
dx = s = x + 1, ds = dx

I =∫
(s − 1)e s −1 ds = e −1 ⎡
es es ⎤
• ⎢∫ s ds − ∫ s 2 ds ⎥⎦
s2 ⎣
Integration by parts on first integral with
1 ds
u = , du = − 2 ; dv = e s ds, v = e s ,
s s
to produce,
−1 ⎡ e es ⎤
s
es
I = e ⎢ + ∫ 2 ds − ∫ 2 ds ⎥
⎣s s s ⎦
⎛ es ⎞ e x +1 ex
= e −1 ⎜⎜ ⎟⎟ = e −1 = .
⎝ s ⎠ x +1 x +1
¾ NOTE: Advanced solution. In first integral above,
es
∫ s ds = Ei(s ) , from definition,
es
∫ s 2 ds = Γ(− 1,−s ) by Formula 2.325.7, above,
(− 1) z +1 a z Γ(− z,−ax n ) (− 1) z +1 a z ∞ 1 −t
n
e ax m −1
∫ xm
dx =
n
=
n ∫ e dt , z =
t z +1 n
− ax n
s
e
Then, Γ(− 1,− s ) = Ei(s ) − by Γ(− 1,− x ) in Sect. 8.352,
s
Sect. 2.32
 58

below so that,

I =∫
(s − 1)e s −1 ds = e −1 ⎡ e s ds − e s ds ⎤
⎢∫ ∫ s 2 ⎥⎦
s2 ⎣ s
⎡ es ⎤ es ex
= e −1 ⎢Ei(s ) − Ei(s ) + ⎥ = e −1 = .
⎣ s ⎦ s x + 1
¾ Cited Dwight, 1961, ITEM 570
xe ax e ax xe x
∫ (1 + ax )2 a 2 (1 + ax )
dx = Exercise. Similar to above, ∫
(1 + x )2
dx .

Check solution on mathematica.com, and compare with

above integral for a = 1.
¾ Cited Dwight, 1961, ITEM 570.1
We show that this integral is equal to 2π for limits ± ∞.
x2
−
∫e 2
dx
It is the basis for the Gaussian probability density
function.
• From Formula 2.33.16, above,

∫ e dx = 2 β erf [ β x]. Set β = 2 and use property for

− βx 21 π 1

error function, Φ (± ∞ ) = ±1 (see Sect. 8.359, below).

Thus,
+∞ 1
⎡1 ⎛ x ⎞⎤ + ∞ 1
2π [erf (+ ∞ ) − erf (− ∞ )]
− x2
∫e
−∞
dx = ⎢
2

⎣2
2π erf ⎜⎜ ⎟⎟⎥
⎝ 2 ⎠⎦ − ∞ 2
=

2π [1 − (− 1) )] = 2π (2 ) = 2π .
1 1
=
2 2
¾ Compare solution with standard approach involving a
double integral transformed to polar coordinates. See
Bers, Calculus, Vol. 2, p. 928. Also shown in books on
probability theory.
x2
¾ NOTE: Since the integral ∫ e − 2 dx contains an even
function, f ( − x) = f ( x), (see GR, Sect. 3.03), then,
+∞ 1 +∞ 1
− x2 − x2
∫e
−∞
2
dx =2 ∫ e
0
2
dx .

In this approach, first solve the indefinite integral,

⎛ 1 x2 ⎞
Γ⎜⎜ , ⎟⎟
−
x2
⎝2 2 ⎠ by Formula 2.33.10,
∫e 2
dx = − 1
⎛1⎞ 2
2⎜ ⎟
⎝2⎠
(
Γ γ , βx n 1 ) ∞
m +1
∫x ∫t
− βx n γ −1 −t
m
e dx = − γ
=− γ e dt , γ = ,
nβ nβ βx n
n
so that,

Sect. 2.32
 59

+∞
⎡ ⎛ 1 x2 ⎞ ⎤
+∞ 1 +∞ 1 ⎢ 2Γ⎜⎜ , ⎟⎟ ⎥
− x2 − x2
⎢ ⎝ 2 2 ⎠⎥
∫ e 2 dx =2 e 2
∫
dx = 2⎢−
2 ⎥
−∞ 0 ⎢ ⎥
⎢⎣ ⎥⎦
0
⎡ ⎛ 1 ⎞ ⎛ 1 ⎞⎤ ⎛1⎞
= 2 ⎢Γ⎜ ,0 ⎟ − Γ⎜ , ∞ ⎟⎥ = 2Γ⎜ ⎟ = 2π
⎣ ⎝ 2 ⎠ ⎝ 2 ⎠⎦ ⎝2⎠

by the properties of incomplete gamma functions, and use

⎛1⎞
of known relation, Γ⎜ ⎟ = π .
⎝2⎠
+∞ 1
− x2
NOTE: In the integral ∫
−∞
xe 2 dx, the integrand is

considered an odd function, f ( − x) = − f ( x), (see GR, Sect.

3.03), so that
+∞ 1
− x2
∫
−∞
xe 2 dx = 0.

+∞ 1
1 − x2
The integral,
2π ∫
−∞
xe 2 dx = 0, is the “mean” of the

Gaussian probability density.

ax See series expansion in Dwight, 1961, ITEM 568.1
e
∫ dx = Ei(ax ) NOTE: Dwight book is useful for other series expansions,
x
and contains some information on the gamma function,
Γ( x ).
xn • This is an important limit for a fundamental
lim = 0 for all values of n relation of the gamma function,
x → +∞ e x
∞
Γ(x + 1) = ∫ t x e −t dt = xΓ(x ). Using integration by parts
0

with u = t , dv = e −t dt , then,
x

t x t = +∞
∞ ∞
Γ(x + 1) = ∫ t e dt = − t
x −t
+ x ∫ t x −1e − t dt
0
e t = 0 0

⎡ tx ⎤
= − ⎢ lim t − 0⎥ + xΓ( x )
⎣t → +∞ e ⎦
= xΓ( x ).
tx
• The limit lim = 0 follows from the theorem which
t → +∞ e t

⎛∞ ⎞
is derived from l'Hôpital's rule ⎜ case ⎟ by taking
⎝∞ ⎠
repeated derivatives of numerator and denominator (see
Bers, Calculus, pp. 547 ff.).

Sect. 2.32
 60

xn
o Try to prove that lim =0
x → +∞ e x

by showing that the numerator eventually becomes n! by

repeated differentiation. For example,
dx 2 2dx
= = 2 ⋅1 = 2!
dx dx
dx 3 3dx 2 dx
= = 3 ⋅ 2 = 3 ⋅ 2 ⋅1 = 3!
dx dx dx
n −1
dx n
dx dx n−2
=n = n(n − 1)
dx dx dx
n −3
dx
= n(n − 1)(n − 2)
dx
dx
= n(n − 1)(n − 2)(n − 3)K3 ⋅ 2
dx
= n(n − 1)(n − 2)(n − 3)K3 ⋅ 2 ⋅1 = n!
tx 1
¾ Apply this result to lim t
= x! lim t = 0.
t → +∞ e t →+∞ e

Sect. 2.32
61
 62

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
Sect. 2.711 • Alternative restatement of 2.711 (2nd formula)
• Derived above in Sect. 2.32 under Formula
m

∫ ln xdx = x∑ (− 1)
m! 2.33.5,
ln m−k x
m k

k =0 (m − k )! ∫ x
n
m ax
e dx =
e ax ⎡ γ −1
⎢
n

∑ (− 1)k (γ − 1)! (x n )γ − k −1 ⎤
⎥
[m ≥ 0] n ⎢⎣ k = 0 (γ − k − 1)! a k +1 ⎥⎦
⎡ m +1 ⎤
⎢γ = n = 1,2,..., n ≠ 0, a ≠ 0⎥
⎣ ⎦
NOTE: unlike 2.711, if m = 0, I = x, a trivial result.
NOTE on notation: (ln x ) = ln m x ≠ ln x m
m

∫ ln (a + bx )dx = • Change of variable,

m

s = ln(a + bx ), ds =
b
a + bx m dx, e s = a + bx, to
∑ (− 1)k m! ln m−k (a + bx ) a + bx
b k =0 (m − k )! give,
es m
I = ∫ s m e s ds = ∑ (− 1)
1 m!
sm−k ,
k

b b k =0 (m − k )!
by above integer-expansion Formula in Sect.
2.32,
eax ⎡ γ −1 (γ −1)! (xn )γ −k −1 ⎤ ⎡γ = m + 1 = 1,2,...⎤
n

∫ = ⎢ ∑ (− )
n
⎥
m ax k
x e dx 1
n ⎣⎢ k =0 (γ − k −1)! ak +1 ⎦⎥ ⎢⎣ n ⎥
⎦
with parameters,
[s = x = ln(a + bx ), m = m, a = n = 1, γ = m + 1]
¾ Reduces to Formula 2.711 for a = 0, b = 1.
Sect. 2.721 Change of variable:
For m = −1 ds 1 s
s = ln x , = , e = e ln x = x.
∫ ln x = Ei(ln x ) = li(x )
dx dx x
es
ds = ∫ ds = Ei (ln x ) = li ( x )
x
I = ∫ s s
by integral previously submitted, Form. 2.325.7,
e ax
n
Ei ax n ( )
∫ x dx =
n
, and Formula 8.240.1.

Sect. 2.721 A standard textbook integral, from

For m = n = 1 Formula 2.721.1 with m = n = 1.
x2 x2
∫ x ln xdx = 2 ln x − 4

—Sect. 2.71 & 2.72-2.73—

 63

Sect. 2.722 • Alternative statement of 2.722

m
m! ln m−k x • Derived above in Sect. 2.32, under
∫ x n
ln m
xdx = x n+1
∑ (− 1) k

k =0 (m − k )! (n + 1)k +1 Formula 2.33.5,

e ax ⎡ γ −1 (γ − 1)! (x n )γ − k −1 ⎤
[m ≥ 0]
n

• ∫ = ∑ (− )
n
⎢ ⎥
m ax k
x e dx 1
n ⎢⎣ k = 0 (γ − k − 1)! a k +1 ⎥⎦
⎡ m +1 ⎤
⎢γ = n = 1,2,..., n ≠ 0, a ≠ 0⎥
⎣ ⎦
x n+1
NOTE: unlike 2.722, if m = 0, I = ,a
n +1
trivial result.

Sect 2.721 or new Section • Change of variable:

∫ ln(ln x ) dx = x ln(ln x ) − nli( x)
n n
dx
s = ln x, ds = , e s = eln x = x . Then,
x
I = ∫ x ln s = ∫ e (n ln s )ds
n s

by relation log x n = n log x

• Integration by parts:
n
u = n ln s, du = ds
s
dv = e s ds, v = ∫ e s ds = e s
es
uv − ∫ vdu = e s (n ln s ) − n∫ ds
s
Substituting s = ln x, ds = , e s = x, n ln s = ln(ln x ) ,
dx n

x
I = x ln(ln x ) − nEi(s ) = x ln(ln x ) − nEi(ln x ) =
n n

x ln(ln x ) − nli(x ) by 8.240.1.

n

¾ Verified Mathematica

—Sect. 2.71 & 2.72-2.73—

 64

Sect 2.721 or new section • Math.com calculations; same

x n +1 ln (ln x ) − mli( x n +1 )
m
approach as for above integral,
∫ x ln(ln x ) dx =
n m

∫ ln(ln x ) dx = x ln(ln x ) − nli( x)

n n
n +1
• For first integral, change of variable:
For n = −1 dx
ln (ln x ) s = ln x, ds = , e s = e ln x = x, x n+1 = e (n+1)s
[ ]
m

∫ x dx = ln x ln(ln x ) − m x
m
n +1
I = ∫x (ln s )m ds = ∫ e (n+1)s ln s m ds = ∫ e (n+1)s (m ln s )ds
by relation log x n = n log x
• Then, int. by parts with,
u = m ln s, dv = e (n+1)s ds.
NOTE: for n = 0, m = n, see above
integral,
∫ ln(ln x ) dx = x ln(ln x ) − nli( x)
n n

• For n = −1 : change of variable &

int. by parts;
dx
s = ln x, ds =
x
m
x ln(s )
I =∫ ds = ∫ m ln sds.
x
int. by parts :
m
u = m ln s,du = ds
s
dv = ds, v = ∫ ds = s

uv − ∫ vds = s (m ln s ) − m ∫ ds =
s
s
ln x ln (ln x ) − m ∫ = ln x ln (ln x ) − m ln x
m dx m

—Sect. 2.71 & 2.72-2.73—

 65

Section 2.724 • Change of variable,

dx
s = ln x, ds = , e s = x, x n +1 = e (n +1)s , to give:
x n dx x
∫ (ln x )m = n
x xds e(n +1)s
⎡m − 2 (m − k − 2 )! (n + 1)k ⎤ (n + 1)m −1 I = ∫ m = ∫ m ds
− x n +1 ⎢ ∑ m − k −1 ⎥
+ ( )
li x n +1 s s
⎣ k = 0 (m − 1)! (ln x ) ⎦ (m − 1)! • In above integral, Sect. 2.32,
− e ax ⎡ z −1 ( z − k − 1)! a k ⎤ Ei(ax n )
n n
e ax
n +1 ∫ dx = ⎢∑ + az
( z −k ) ⎥ ,
( )
n m
x n ⎣ k =0 z! ⎦ n z!
∫ (ln x )2 = − ln x + (n + 1)li x
x dx x n +1 xn
set s = x = ln x,n = 1,a = n + 1,z = m − 1

n +1 ⎡ 1 ⎤ (n + 1) e(n +1)s
( ) and substitute for solution of I = ∫
2
x n dx 1
∫ (ln x )3 = − x ⎢⎣ 2(ln x )2 + 2 ln x ⎥⎦ + 2 li x
n +1 ds .
sm
• Alternative statement of solution:
x n dx ⎡m −2 (n + 1)m−k −2 ⎤
∫ (ln x )m ⎢∑
n +1
= − x k +1 ⎥
⎣ k =0 (m − 1)(m − 2 )K (k + 1)(ln x ) ⎦

+
(n + 1)m−1 li(x n+1 )
(m − 1)!

e(n +1)s
based on converting I = ∫ ds into gamma
sm
n
e ax
function via Formula 2.325.8, ∫ m dx,
x
to give, (− 1) (n + 1) Γ[− m + 1,−(n − 1)ln x ] , and
m m −1

using Formula 8.352 5, Γ(− n + 1, x ) , which

produces a different ordering of the series terms.
NOTE: 2.724.2 follows if m = 1.
NOTE: m = 2,3 : substitute into expansion formula
Formula 2.724.2 ds 1 s
Set s = ln x, = , e = e ln x = x.
dx x
Change Formula 2.724.2 to read: Then,

I =∫
xx n
ds = ∫
es ( ) n +1

ds = Ei[(n + 1)s ]
( )
n
= Ei[(n + 1) ln x ] = li x n+1
x dx
∫ s s
ln x
( ) ( )
= Ei[(n + 1) ln x ] = li e (n+1)ln x = li x n+1 ,
by Formula 2.325 7,
e ax
n
( )
Ei ax n
( )
∫ x dx = n , Formula 8.240.1, Ei(x ) = li e ,
x

and relation e x ln a = a x .
NOTE: new version relates exponential- and
logarithm-integrals, Ei( x ) & li( x ), Formula 8.211.1.

—Sect. 2.71 & 2.72-2.73—

 66

New Section • Integration by parts:

∫ (a + bx ) ln (c + ln x )dx = ⎛1⎞
m n
n⎜ ⎟ ln n −1 (c + ln x )
⎡(a + bx )m +1 ln n (c + ln x ) − u = ln (c + ln x ), du = ⎝ ⎠
n x
⎤ dx
1 ⎢ ⎥ c + ln x
⎢
(m + 1)b ⎢n ∫ (a + bx ) ln n −1 (c + ln x ) ⎥
m +1
s = a + bx 1 (a + bx )m+1
dv = (a + bx ) dx, v =
dx ⎥
b∫
x(c + ln x ) =
m m
⎣ ⎦ s ds
b(m + 1)
For m = −1 uv − ∫ vdu =
ln n (c + ln x ) ln(a + bx ) ln n (c + ln x ) ⎡(a + bx )m +1 ln n (c + ln x ) − ⎤
∫ a + bx dx = −
1 ⎢ ⎥
b
⎢
(m + 1)b ⎢n ∫ (a + bx ) ln n −1 (c + ln x ) ⎥
m +1

n ln (a + bx ) ln n −1 (c + ln x ) dx ⎥
x(c + ln x )
b∫
dx ⎣ ⎦
x(c + ln x )
• For m = −1 :
Integration by parts with,
u = ln n (c + ln x ), dv =
dx
.
a + bx
NOTE: a difficult (to reduce further) but general
integral useful for deriving existing forms by setting
a, b, c, m, n as desired.
New Section • Integration by parts:
nk ln n −1 (c + kx )
u = ln n (c + kx ), du =
∫ (a + bx ) ln n (c + kx )dx =
m dx
c + kx
s = a + bx 1
⎡(a + bx )m +1 ln n (c + kx ) − ⎤ dv = (a + bx ) dx, v = ∫ (a + bx ) dx =
b∫
s m ds =
m m

1 ⎢ ⎥
( ) ( )
m +1 n −1
⎢
(m + 1)b ⎢nk + + ⎥
∫
a bx ln c kx
dx ⎥ (a + bx )m+1
⎣ c + kx ⎦ b(m + 1)
For m = −1 uv − ∫ vdu
ln(c + kx )
n

∫ dx = =
1
(a + bx )m+1 ln n (c + kx )
a + bx (m + 1)b
⎡ln (a + bx ) ln n (c + kx ) − ⎤
1⎢ ⎥ nk (a + bx ) ln n −1 (c + kx )
m +1

b(m + 1) ∫
⎢ ln (a + bx ) ln
n −1
(c + kx ) ⎥ − dx
c + kx
⎢⎣ ∫
b nk dx ⎥
c + kx ⎦
• NOTE: reduces to 2.278.1 if
a = 0, b = 1, m = m, c = a, k = b, n = 1.
• Test: if
a = c = m = 0, n = b = k = 1, I = ∫ ln xdx = x ln x − x.
• For m = −1 :
dx
Integration by parts: u = ln(c + kx), dv =
a + bx
• Reduces to 2.721.2 if c = a = 0, b = k = 1.
• NOTE: a difficult (to reduce further) but general
integral useful for deriving existing forms (such
as 2.711, 2,722, 2.733, 2.726) by setting
a, b, c, k , m, n as desired.
—Sect. 2.71 & 2.72-2.73—
 67

Sect. 2.727 or New Section • Integration by parts:

(ln x )n dx = u = (ln x ) , dv = (a + bx ) dx
n −m

∫ (a + bx )m • NOTE: 2.727.1 results if n = 1.

• NOTE: may be derived from
⎡
−
(ln x )n + ⎤ (a + bx )m ln n (c + kx )dx =
⎢ ⎥
⎢ (a + bx )
m −1
(a + bx )m +1 ln n (c + kx ) − nk ∫ (a + bx ) ln (c + kx ) dx⎥
⎡ ⎤
⎥
m +1
1 ∫ 1 n −1

(m + 1)b ⎢⎣
b(m − 1) ⎢
c + kx
(ln x ) n −1 ⎥ ⎦
⎢n ∫ dx ⎥ if c = 0, k = 1, m = −m.
⎢⎣ x(a + bx )
m −1
⎥⎦ • m =1:
Integration by parts with u = (ln x ) , dv =
n dx
For m = 1 .
a + bx

(ln x )n dx =
∫ a + bx
(ln x )n ln(a + bx) − n (ln x)n−1 ln(a + bx) dx
b b∫ x
New Section • First integral:
(ln x ) dx = 1 ⎡− (ln x ) + n (ln x ) dx ⎤
n n n −1 Set a = 0, b = 1 in above integral,
∫ xm ⎢
m − 1 ⎣ x m −1
∫ m ⎥ (ln x )n dx = 1 ⎡− (ln x )n + n (ln x )n−1 dx⎤
x ⎦ ∫ ⎢ ∫ ⎥
m
(a + bx) m−1
b(m − 1) ⎣
m−1
(a + bx ) x(a + bx ) ⎦
For m = 1, see 2.721.2 • For m = 1 : see 2.721.2 with m → n.
• Expansion formula:
(ln x )n dx = −1 ⎡ n n! (ln x ) ⎤
n−k ™ Change of variable;
∫ xm ⎢∑ ⎥ dx
x m −1 ⎣ k = 0 (n − k )! (m − 1)k +1 ⎦ s = ln x, ds = , e s = x, x1− m = e(1− m )s
x
ln x 1 ⎡ ln x 1 ⎤ n
∫ x m dx = − x m −1 ⎢⎣ m − 1 + (m − 1)2 ⎥⎦ s
I = ∫ m xds = ∫ s n e(1− m )s .
x
™ Now use indefinite integral previously
(ln x )2 dx = derived (omitted in GR 7, but submitted as
∫ xm
erratum),

1 ⎡ (ln x )
(− 1) Γ γ ,−ax n
1−γ
(m +1 )
2 ⎤ ∫ x e dx = , γ =
2 m ax n
2 ln x γ
− m −1 ⎢ + + ⎥ na n
x ⎣ m − 1 (m − 1) (m − 1)3 ⎦
2
with parameters [a = 1 − m, γ = n + 1] to give:
(− 1)− n Γ[n + 1,−(1 − m)ln x]
(ln x )3
I=
∫ dx = (1 − m )n+1
xm
Γ[n + 1, (m − 1)ln x ]
⎡ (ln x )3 3(ln x )2 6 ln x ⎤ =−
1
⎢ + +
m − 1 (m − 1) (m − 1)3 ⎥
2 ⎥ (m − 1)n+1
− m−1 ⎢ ⎡ m!e −(m−1)ln x n [(m − 1)ln x ]k ⎤
x ⎢ ⎥
n +1 ∑
⎢+
6
⎥ = −⎢ ⎥
⎣ (m − 1) ⎣ (m − 1) k! ⎦
4
⎦ k =0

⎡ m! 1 n [(m − 1)ln x ] ⎤
k

m −1 ∑
For m = 1 and n = −1 , see 2.721.3 = −⎢ ⎥ by 8.352.2,
⎣ (m − 1) x k =0
n +1
k! ⎦
Γ(n + 1,x ).
™ Writing summation in reverse order:
—Sect. 2.71 & 2.72-2.73—
 68

(ln x)n dx = − 1 ⎡ ln x n n(ln x )

n −1
n(n − 1)(ln x )
n−2
n! ⎤
∫ xm
⎢ +
x ⎣ m − 1 (m − 1)
m−1 2
+
(m − 1) 3
+K+
(m − 1)n+1 ⎥⎦

1 ⎡ n n! (ln x ) ⎤
n−k
=− ⎢∑ ⎥ by inspection.
x m −1 ⎣ k = 0 (n − k )! (m − 1)k +1 ⎦
NOTE: this formula may be derived readily by
restating
( ln x )
n
I = ∫ m dx = ∫ s n e − (m −1)s ds and applying integral
x
derived above in Sect. 2.32,
− e − βx ⎡ γ −1 (γ − 1)! (x n )
γ − k −1
n
⎤ ⎡ m +1 ⎤
∫
m − βx
= ∑ ⎥ ⎢γ = = 1,2,...⎥
n
x e dx ⎢
n ⎢⎣ k = 0 (γ − k − 1)! β k +1
⎥⎦ ⎣ n ⎦
with parameters
[ n = 1, m → n, γ = n + 1, β x n = e − (m −1)s , s = x = ln x. ]
• For n = 1,2,3 : substitute into expansion formula
above
• For m = 1 and n = −1 , see 2.721.3.
Sect. 2.728 • New reduction formula; same reduction
approach as Formula 2.728.1
∫x
m
ln n (a + bx )dx = • Integration by parts:
⎡ x m+1 ln (a + bx )n − ln n −1 (a + bx )
1 ⎢
⎤ u = ln n (a + bx ), du = nb dx,
⎥ a + bx
⎢
m + 1 nb x m +1
ln n −1
(a + bx ) dx ⎥⎥
⎣⎢ ∫ a + bx ⎦ dv = x m dx, v = ∫ x m dx =
x m +1
;
m +1
For n = −1 uv − ∫ vdu
dx li(a + bx )
∫ ln(a + bx ) = b x m +1 n
ln (a + bx ) −
nb x m +1 ln n −1 (a + bx )
m +1∫
= dx
m +1 a + bx
For n = 1, m = −1 , see 2.728.2 . Also, see • n = −1 :
o Change of variable:
Sect. 2.32, above, for multiple solutions to
u = a + bx, du = bdx
ln (a + bx )
∫ x dx (in terms of dilogarithms, I= ∫
1 du Ei(ln u ) li(u ) li(a + bx )
= = =
defined in Math. Summary) all of which b ln u b b b
differ from equivalent GR solution! by Formula 8.240.1.
¾ Verified math.com
x • Change of variable:
∫ ln(a + bx ) dx = s−a
s = a + bx, ds = bdx, x =
Ei[2ln (a + bx )] − aEi[ln (a + bx )] b
= 1 (s − a ) / b 1 ⎡ s ds ⎤
b2 I= ∫ ds = 2 ⎢ ∫ ds − a ∫
[ 2
]
li (a + bx ) − ali(a + bx ) b ln s
s
b ⎣ ln s ln s ⎥⎦
b2 o ∫ ds :
ln s

—Sect. 2.71 & 2.72-2.73—

 69

du 1 u
u = ln s, = ,e = s
ds s
s2 e 2u
I = ∫ du = ∫ ds =Ei(2u ) =
u u
[
Ei(2 ln s ) = Ei[2 ln (a + bx )] = li (a + bx )
2
]
by 8.211.1 and relation e x ln a = a x .
ds
o a∫ :
ln s
du 1 u
u = ln s, = ,e = s
ds s
eu
I = a ∫ ds = a ∫ du = aEi(u ) =
s
u u
aEi(ln s ) = aEi[ln(a + bx )] = ali(a + bx )
by 8.211.1 and relation e x ln a = a x .
xdx Ei[2ln (a + bx )] − aEi[ln (a + bx )]
• ∫ =
ln(a + bx ) b2
NOTE: if
a = 0, b = 1, I = ∫
x
ln x
( )
dx = li x 2 by 2.724 2.
¾ verified math.com
New Section • Change of variable:
ln m (a + ln x ) dx
∫ x n dx = s = a + ln x, ds =
x
ln m (a + ln x ) m ln m −1 (a + ln x )
a + ln x
e =e
s
= xe ⇒ x = e s e −a
a

(n − 1)x n −1 n − 1 ∫ x n (a + ln x )
− + dx
e (1−n )s
x1−n =
e (1−n )a
For n = 1
ln m (a + ln x ) x(ln s ) ds
m

∫ x
dx = I =∫
x n
= ∫ x1− n (ln s ) ds =
m

(a + ln x )ln m (a + ln x ) k =1− n 1
1
(1− n )a ∫ e (1− n )s
(ln s )m
ds = ka ∫ e ks (ln s ) ds.
m

ln m −1 (a + ln x ) e e
− m∫ dx
• Integration by parts on ∫ e ks (ln s ) ds :
m
x
m(ln s )
m −1
For m = 1 and n = 1 u = (ln s ) , du =
m
ds
s
See integral below, 1
dv = e ks ds, v = ∫ e ks ds = e ks
ln (a + ln x ) k
∫ x dx = e (ln s )
ks m
m e ks (ln s ) ds
m −1

(a + ln x ) ln(a + ln x ) − ln x uv − vdu = − ∫
k k s
Substituting

—Sect. 2.71 & 2.72-2.73—

 70

dx
s = a + ln x, ds = , k = 1 − n, e(1− n )s = x1− n e(1− n )a ,
x
1− n (1− n )a
⎡x e ln m (a + ln x ) ⎤
1 ⎢ 1− n ⎥
I= ⎢ ⎥=
e (1− n )a
⎢ m x1− n e(1− n )a ln m −1 (a + ln x ) ⎥
⎢− 1 − n ∫ x(a + ln x )
dx ⎥
⎣ ⎦
ln m (a + ln x ) m ln m −1 (a + ln x )
(n − 1)x n −1 n − 1 ∫ x n (a + ln x )
− + dx.

NOTE: easier approach: int. by parts on

u = ln m (a + ln x ), dv = n
dx
x

• For n = 1 :
ln m (a + ln x )
∫ x
dx =

(a + ln x ) ln m (a + ln x ) − m∫ ln (a + ln x ) dx
m −1

x
derived by change of variable with
dx
s = a + ln x, ds = , to give,
x
I = ∫ (ln s ) ds, and integration by parts with :
m

u = (ln s ) , dv = ds.
m

¾ Mathematica provided solution in terms of

hypergeometric function.

• Similar to ∫ ln (a +n ln x )dx above: make n → −n.

m
New Section
∫x
n
ln m (a + ln x )dx = x
• For m =1:
⎡ x n+1 ln m (a + ln x ) ⎤ o Change of variable:
1 ⎢ ⎥
n + 1 ⎢− m x ln (a + ln x ) dx ⎥
n m −1
dx
⎢⎣ ∫ a + ln x ⎥⎦ s = a + ln x, ds =
x
a + ln x
e =e
s
= xe ⇒ x = e s e − a
a
For m = 1
e (n +1)s k = n +1 e ks
∫ x ln(a + ln x )dx =
n
x n +1 = = ka
e (n +1)a e
1 ⎪⎧ x ln(a + ln x ) − ⎫⎪
n +1

⎨ −(n +1)a ⎬ 1
n + 1 ⎪⎩e Ei[(n + 1)(a + ln x )]⎪⎭ I = ∫ x n +1 ln s = ka ∫ e ks ln sds
e
o Integration by parts:

For m = 1and n = −1
ln (a + ln x )
∫ x dx =
(a + ln x ) ln(a + ln x ) − ln x

—Sect. 2.71 & 2.72-2.73—

 71

ds
u = ln s, du =
s
1
dv = e ks ds, v = e ks
k
ln e ns 1 e ks
ks
uv − vdu = − ∫ ds =
k k s
ln e ks ns Ei(ks )
− .
k k
Substituting
dx
s = a + ln x, ds = , k = n + 1, e (n +1)s = x n +1e (n +1)a ,
x
⎡x en +1 ( n +1)a
ln (a + ln x ) ⎤
−⎥
1 ⎢ n +1
I = (n +1)a ⎢ ⎥
e ⎢ 1 ⎥
⎢⎣ n + 1 Ei(n + 1)(a + ln x ) ⎥⎦

=
1
n +1
[ ( )]
x n +1 ln (a + ln x ) − e −(n +1)a li x n +1e (n +1)a

( )
by Formula 8.211.1, Ei( x ) = li e x , and relation,
e x ln a = a x .
• For m = 1 and n = −1 : math.com calculation;
(change of variable,
s = a + ln x and int. by parts on u = ln s,dv = ds )

New Section
∫ x ln(a − ln x )dx = • Similar to ∫ x n ln(a + ln x )dx , above
n

1 ⎧⎪ x ln(a − ln x ) −
n +1
⎫⎪ • n = −1 :
⎨ −(n +1)a ⎬ ln (a + ln x )
Ei[(n + 1)(ln x − a )]⎪⎭
n + 1 ⎪⎩e Similar to ∫ x
dx , above

For n = −1
ln(a − ln x )
∫ x dx = (ln x − a )ln(a − ln x ) − ln x
New Section • Change of variable:
dx dx
∫ (a − ln x )n = s = a − ln x, ds = −
x
1
x 1 dx ln ea
− ∫
a − ln x
e =e
s
=e e =
a x
(n − 1)(a − ln x ) n −1
n − 1 (a − ln x )n −1 x
a −s −s −a
x = e e , e = xe
For n = 1 −s
x a e
I = −∫ ds = − e ∫ s n ds
∫ a − ln x = −e Ei(ln x − a )
dx a
sn
• Integration by parts:

—Sect. 2.71 & 2.72-2.73—

 72

dx u = e − s , du = −e − s ds
∫ (a − ln x ) n
=
s 1− n 1
dv = s ds, v = ∫ s ds =
−n −n
=−
n−2
(− 1)n x∑ (− 1)k (n − k − 2)! +
1− n (n − 1)s n−1
k =0 (n − 1)!(a − ln x )n−k −1 uv − ∫ vdu =
a
(− 1)n e
Ei(ln x − a ) e −s − e −s
(n − 1)! (n − 1)s n−1 ∫ − (n − 1)s n−1
− − ds

e −s e −s s = a − ln x

(n − 1)s n−1 ∫ (n − 1)s n−1

=− − =
∫ (a − ln x )2 = a − ln x + e Ei(ln x − a )
dx x a ds

xe − a 1 − xe − a
n − 1 ∫ x(a − ln x )n −1
− − dx.
(n − 1)(a − ln x )n−1
⎡ xe − a ⎤
⎢ − ⎥
a ⎢ (n − 1)(a − ln x )
n −1

I = −e ⎥
⎢ 1 − xe −a ⎥
⎢− ∫
⎢⎣ n − 1 x(a − ln x )
n −1
dx ⎥
⎥⎦
x 1 dx
=
(n − 1)(a − ln x ) n −1
− ∫
n − 1 (a − ln x )n −1
• For n = 1 : Change of variable;
1
dx s a − ln x
ln ea
s = a − ln x, ds = − , e = e =e e x =
a
⇒
x x
x = e a e −s .
e −s
I = −e a ∫ ds = −e a Ei(− s ) = − e a Ei(ln x − a )
s
by 8.211.1 and relation, e x ln a = a x .
• Integer-expansion formula:
From above,
e− s
I = −e a ∫ n ds
s
Apply integral formula, above, Sect. 2.32,
e−βx ( −1) e−βx ⎡ z−1 k ( z − k − 1)! β ⎤
n z n
k

∫x m
dx = ⎢∑ (− 1) ( z −k ) ⎥
n ⎣ k =0 z! xn ⎦

+ (−1) β z
(
Ei − βxn ) ⎡ m-1 ⎤
⎢⎣ z = n = 1,2,..., m = 2,3,...⎥⎦
z

n z!

with parameters,
[s = x = a − ln x, n = 1, m = n, β = 1, z = n − 1]
o Alternative expression of expansion
formula:

—Sect. 2.71 & 2.72-2.73—

 73

dx
∫ (a − ln x ) n
=

(− 1) x∑
n
n−2
(− 1)
k
+
k = 0 (n − 1)(n − 2 )K (k + 1)(a − ln x )
k +1

(− 1)n ea Ei(ln x − a )
(n − 1)!
e−βx
n

based on expansion formula for ∫ m dx previously

x
submitted as Form. 2.325.19, which produces a
different ordering of the series terms.
• n = 2 : In above integral,
dx x 1 dx
∫ (a − ln x )n = (n − 1)(a − ln x )n−1 − n − 1∫ (a − ln x )n−1
,
set n = 2, and note,
dx s = a −ln x a e − s
∫ a − ln x = − e ∫ s ds = −e Ei(− s ) =
a

− e a Ei(ln x − a )
NOTE: Formula 4.212.4 results for this integral
with limits from 0 to 1.
New Section • First integral:
o Change of variable:
dx dx
∫ x (ln x )
n m
= s = ln x, ds = , e s = x, x1− n = e(1− n )s to give,
x
1 n −1 dx x e(1− n )s e − (n −1)s
− − ∫ I = ∫ n m ds = ∫ m ds = ∫ m ds
(m − 1)x (ln x )
n −1 m −1
m − 1 x (ln x )m −1
n
x s s s
o Integration by parts:
For m = 1 u = e (1−n )s , du = (1 − n)e (1−n )s
dx ⎛ 1 ⎞ 1
∫ x n ln x = li⎜⎝ x n−1 ⎟⎠ dv = s −m ds, v = ∫ s −m ds = −
(m − 1)s m−1
e (1−n)s − (n − 1) e (1−n )s
uv − ∫ vdu = −
(m − 1)s m−1 − (m − 1) ∫ s m−1
o − ds
For n = 1
x1−n (n − 1) x1−n dx
(m − 1)(ln x)m−1 (m − 1) ∫ (ln x)m−1 x
dx 1 =− −
∫ x(ln x )m = − (m − 1)(ln x )m −1
For n = m = 1, see 2.721 .3 1 (n − 1) dx
=−
(m − 1)x (ln x)
n−1 m−1
− ∫
(m − 1) x (ln x)m−1
n

NOTE: setting m → −m, n → −n in 2.721.1

produces a poor reduction formula since powers of
ln x are increasing. Setting n → − n in 2.724
produces desired results.

• For m = 1:

—Sect. 2.71 & 2.72-2.73—

 74

dx dx s
∫ x (ln x )
n m
= s = ln x, ds = , e = x, x1− n = e(1− n )s , to give
x
e (1−n )s
[ ]
1− n
k (m − k − 2 )! (n − 1) ds =Ei[(1 − n ) ln x ] = li e (1−n )ln x
1 ⎡m − 2 ⎤ x
I =∫ ds = ∫
k
− n −1 ⎢ ∑ (− 1) m − k −1 ⎥
+
x ⎣ k =0 (m − 1)! (ln x ) ⎦ s s

(n − 1)
m −1
⎛ 1 ⎞ [ ] ⎡ 1 ⎤
= li x1−n = li ⎢ n−1 ⎥
(− 1) m −1
li⎜ ⎟ ⎣x ⎦
(m − 1)! ⎝ x n −1
⎠
[n > 1] o NOTE: it is unwise to set n = 1 in this
formula to solve Formula 2.721.3,
dx n=1
∫ x ln x = li(1).
• For n = 1 :
dx 1
∫ x(ln x )m dx = − (m − 1)(ln x )m −1 by change of
variable,
s = ln x.
• Integer-expansion formula:
e − (n−1)s
Define from above, I = ∫ m ds
s
− βx n
e
and use integral, ∫ m dx, Section 2.32, above,
x
k ( z − k − 1)! β
− βx n − βxn
⎡ z−1 k
⎤
e e
∫ xm dx = − n ⎢⎣∑ (−1) ( z −k ) ⎥ +
k =0 z! x n
⎦

(−1)z β z Ei(− βx ) ⎡ m −1 ⎤
n

n z! ⎢⎣z = n = 1,2,...,m = 2,3,...⎥⎦

with parameters
[x = s = ln x, n = 1, β = n − 1, z = m − 1].
• Alternative form for expansion:
dx
∫ x n (ln x )m =
1 ⎡m−2 (− 1)k (n − 1)m−k −2 ⎤
n −1 ⎢ ∑
− ⎥+
x ⎣ k =0 (m − 1)(m − 2)K (k + 1)(ln x )k +1 ⎦

(− 1)m−1 (n − 1)
m −1
⎛ 1 ⎞
li⎜ n −1 ⎟
(m − 1)! ⎝x ⎠
based on first submission for expansion of
e − βx
n

∫ x m dx ,
Formula 2.325.19, which produces a different
ordering of the series terms.
¾ EXAMPLES:
dx s = ln x 1
o ∫ = −
x(ln x )
2
ln x

—Sect. 2.71 & 2.72-2.73—

 75

dx s = ln x ⎛1⎞
o ∫ x 2 ln x = Ei(− ln x ) = li⎜⎝ x ⎟⎠ ; or by
expansion formula above ( n = 2, m = 1)
dx s = ln x 1 ⎛1⎞
o ∫ 2 = − − li⎜ ⎟ ; or by
x (ln x ) ⎝x⎠
2
x ln x
expansion formula above ( n = m = 2).

New Section • First two integrals: Similar to above integrals,

dx dx
dx ∫ (a − ln x )n & ∫ a − ln x .
∫ (a + ln x ) n
=
• Integer-expansion formula
x 1 dx Change of variable:
−
(n − 1)(a + ln x ) n −1
+ ∫
n − 1 (a + ln x )n −1 dx
s = a + ln x, ds = , e s = e a eln x = xe a .
x
For n = 1 s
e ds
I = e− a ∫ n .
∫ a + ln x = e Ei(a + ln x )
dx −a s
Apply above integral, Sect. 2.32,

dx
n
e ax − e ax ⎡ z −1 (z − k − 1)! a k ⎤
n
( )
z Ei ax
n

∫ (a + ln x ) n
= ∫ xm dx = ⎢
n ⎣k =0
∑ z!
⎥
x n( z − k ) ⎦
+ a
n z!
⎡ m −1 ⎤
− x∑
n−2
(n − k − 2)! +
e −a
Ei(a + ln x )
⎢ z = n = 1,2,..., m = 2,3,...⎥
⎣ ⎦
k = 0 (n − 1)!(a + ln x )
n − k −1
(n − 1)! with parameters
[x = s = a + ln x, n = 1, m = n, a = 1, z = n − 1]
o Alternative form for expansion:
n−2
dx 1
dx x
+ e −a Ei(a + ln x ) ∫ (a + ln x )n = − x ∑ +
∫ (a + ln x )
k +1
=− k = 0 ( n −1)( n − 2 ) K ( k +1)( a + ln x )
2
a + ln x
Ei(a + ln x )
e− a
(n − 1)!
n
e ax
based on first submission for expansion of ∫ m dx ,
x
Form. 2.325.8, which produces a different ordering
of the series terms.
• n = 2 : Similar to above integral,

∫ (a − ln x )2 = (a − ln x ) + e Ei(ln x − a ) , by setting
dx x a

n = 2 in,
dx n−2
(n − k − 2)! + e−a Ei(a + ln x )
= − x ∑
∫ (a + ln x)n k =0 (n − 1)!(a + ln x)n−k −1 (n − 1)!
o NOTE: Formula 4.212.3 results for this
integral with limits 0 to 1.

—Sect. 2.71 & 2.72-2.73—

 76

New Section • First integral

xn Change of variable:
∫ (a + ln x )m dx = dx s
s = a + ln x, ds = , e = xe a , x n+1 = e (n+1)s e −(n+1)a ,
x
⎡m − 2 (m − k − 2 )! (n + 1)k ⎤
− x n +1 ⎢ ∑ m − k −1 ⎥
+ to get :
⎣ k =0 (m − 1)! (a + ln x ) ⎦ e (n+1)s
I = e −(n+1)a ∫
e −(n +1) a
(n + 1)m −1
Ei[(n + 1)(a + ln x )] sm
ds
(m − 1)! o Solve I with expansion formula in
Sect. 2.32, above:
e ax ⎡ z −1 (z − k − 1)! a k ⎤ Ei(ax n )
n n
e ax
⎢∑
For m =1:
∫ xm
dx = −
n ⎣k =0 z!
+ az
⎥
x n( z −k ) ⎦ n z!
xn
∫ a + ln x dx = e Ei[(n + 1)(a + ln x )]
−( n +1) a
⎡ m −1 ⎤
⎢n ≠ 0, a ≠ 0, z = n = 1,2,..., m = 2,3,...⎥,
⎣ ⎦
with parameters [a = n + 1, n = 1, m = m, z = m − 1]
xn
∫ a − ln x dx = −e Ei[(n + 1)(ln x − a )]
( n +1) a
• Second & 3rd integrals: substitute m = 1 into
expansion formula & s = a − ln x into 3rd.
¾ Verified Mathematica
ln x math.com calculation.
∫ (a + ln x ) 2
dx =
See derivation of definite integral, Sect. 4.212,
below.
+ (1 − a )e −a Ei(a + ln x )
ax
a + ln x
ln x math.com calculation
∫ (a − ln x )2 dx = See derivation of definite integral, Sect. 4.212,
below.
+ (a + 1)e a Ei(ln x − a )
ax
a − ln x
n
e ax Express exponential integrals as equivalent integrals
∫ x m dx in logarithms.
NOTE: see Sect. 4.215, below, for similar definite
∫ x e dx
m ax n
integrals.

—Sect. 2.71 & 2.72-2.73—

 77

DEFINITE INTEGRALS OF
ELEMENTARY FUNCTIONS
78
79
 80

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
e− pu − e− pv Solve indefinite integral and put limits.
v

∫ e dx =
−ρ x
[Re p > 0] v
e− p x v e− pu − e− pv
∫ e dx = −
−ρ x
u
p = .
u
p u p
NOTE: if u = 0, v = ∞, GR Formula 3.310 results, one of the
most elementary of the elementary exponential integrals in the
GR text.
∞
e− x − e− a x An ingenious approach suggested in 1934 modern analysis book
∫0 x dx = ln a [Re a > 0] by Whittaker/Watson (Sect: Gamma Functions).
∞
1
∫e
− ax
• Solve dx =
, a > 0.
0
a
Then, from definition of logs,
1
a a ∞
⎛ ⎞
ln a = ∫ da = ∫ ⎜⎜ ∫ e −ax dx ⎟⎟da so that
1
a 1⎝0 ⎠
∞ a
⎛ ⎞ ∞
⎛ e −ax a= a ⎞
ln a = ∫ ⎜⎜ ∫ e −ax da ⎟⎟dx = ∫ ⎜⎜ − ⎟dx
a = 1⎟
0⎝1 ⎠ 0⎝
x ⎠
∞
e − x − e −ax
=∫ dx.
0
x
NOTE: By GR 3.434.2 solution (see below) can be found by
differentiation under the integral sign.
¾ EXERCISE: Using above approach, show
∞
exp(− ax) − exp(−bx)
∫ dx = ln
b
[Re a, b > 0]
0
x a
o Compare with approach in Section. 3.434, below.
NOTE: These integrals fall under a general approach called
“Frullani Theorem,” of form,
f (ax ) − f (bx )
∞
dx = ln [ f (0 ) − f (∞ )]
b
∫
0
x a
See Moll’s article at:
http://math.tulane.edu/~vhm/Table.html
NOTE: Is integral GR 3.476.1 conformable to Frullani Theorem
(why?):
∞
exp(− ax p ) − exp(−bx p ) 1 b
∫0 x
dx = ln
p a

——Sect. 3.310——
 81

PROOF OUTLINE
(formula references are to G & R 7e)
0
xe x • Catalan’s constant G; see Table of Integrals, Series and
−∫ dx = G Products (“Use of the Tables” & Sect. 9.73), and
−∞
1 + e2 x
Mathematica functions website.
1
ln x
• In integral, Formula 4.231.12, G = − ∫ dx, use
0
1 + x2
change of variable:
dx
set s = ln x, ds = , s0 = −∞, s1 = 0, e s = x, x 2 = e 2 s .
x
¾ Verified math.com
u
ex
n
ln a u
ax
n
li a u ( )
n
u
ex
n
ln a
1
( ) ( )
Ei u n ln a − 0
∫ ∫ x ∫
u
dx = dx = I= dx = Ei x n ln a = =
−∞
x −∞
n −∞
x n −∞ n
[n > 0] ( n
li e u ln a
=
) ( )
li a u
,
n

n n

by ∫
e ax
n

dx =
Ei e ax
n
( )
x n
¾ Verified math.com

( )
v v In indefinite integral, Sect. 2.312, above,
∫a dx = ∫ exp x n ln a dx
xn

u u
(
∫ exp x ln a dx = ∫ a dx =
n xn
)
⎛1 ⎞ ⎛1 ⎞ 1 ⎛1 ⎞,
Γ⎜ ,−u n ln a ⎟ − Γ⎜ ,−v n ln a ⎟ − Γ⎜ ,− x n ln a ⎟
= ⎝ ⎠ ⎝n ⎠ ⎝n ⎠
1
n
n(− ln a )
n
1

n(− ln a )
n
put limits (u,v).
[Re n > 0, Re ln a < 0] ¾ Verified math.com

For first Integral, set u = 0, v = 1 in above integral,

( )
1 1

∫ a dx = ∫ exp x ln a dx
xn n
⎛1 ⎞ ⎛1 ⎞
Γ⎜ ,−u n ln a ⎟ − Γ⎜ ,−v n ln a ⎟
,
( )
0 0 v v
⎝ n ⎠ ⎝ n ⎠
∫ a dx = ∫ exp x ln a =
xn n

⎛1 ⎞
1

γ ⎜ ,− ln a ⎟ n(− ln a )
n
u u

[Re n > 0, Re ln a < 0]

= ⎝ ⎠ [Re n > 0, Re ln a < 0]
n
1
⎛1 ⎞ ⎛1 ⎞ ⎛1 ⎞
Γ⎜ ,0 ⎟ − Γ⎜ ,− ln a ⎟ γ ⎜ ,− ln a ⎟
n(− ln a )
n
⎝ n ⎠ ⎝ n ⎠= ⎝ n ⎠, noting that
1 1

n(− ln a ) n(− ln a )
n n

Γ(a ) = γ (a, x ) + Γ(a, x ).

• 2nd and 3rd integrals are similar
¾ Verified math.com

——Sect. 3.311—
 82

( )
1 1
1
∫a
0
xn
dx = ∫ exp − x n ln a dx
0

⎛1 ⎞
γ ⎜ , ln a ⎟
= ⎝ ⎠ [Re n > 0, Re ln a > 0]
n
1

n(ln a )
n

⎛1⎞
∞ ∞ Γ⎜ ⎟
1
(
∫0 a xn dx = ∫0 exp − x ln a dx =
n
) ⎝n⎠
1

n(ln a )
n

1 1
⎛ 12 ⎞ In above integral,
∫0 = ∫0 ⎜⎜ x ln a ⎟⎟dx =
x
a dx exp ⎛1⎞ ⎛1 ⎞
⎝ ⎠ Γ⎜ ⎟ − Γ⎜ ,− ln a ⎟
( )
1 1
⎝n⎠ ⎝n ⎠,
2[1 − a (1 − ln a )] ∫0 a dx = ∫0 exp x ln a dx =
xn n

[Re ln a < 0] n(− ln a )

1

(ln a )2
n

1
set n = , to give,
2
I=
2
[1 − Γ(2,− ln a )] =
(− ln a )2
2(a ln a − a + 1)
2
[ ]
1 − e ln a (1 − ln a ) =
(ln a ) 2
(ln a )2
2[1 − a(1 − ln a )]
= ,
(ln a )2
by relation Γ(2,− x ) = e x (1 − x ) , Sect. 8.352, below.
¾ Verified math.com

——Sect. 3.311—
83
 84

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
Formula 3.321 1
After
u In derived integral 3.381.8,
∫ e dx
−x 2

( ) ( )
u
−1
∫x e
m −βxn
0
dx = nβν γ ν , βu n
0
Add m +1
ν= [ u > 0, Reν > 0, Ren > 0, Reβ > 0 ]
1 ⎛1 ⎞ n
= γ ⎜ ,u2 ⎟
2 ⎝2 ⎠ ⎡ 1⎤
set ⎢m = 0, n = 2, β = 1,ν = ⎥
⎣ 2⎦
giving,
1 ⎛1 2⎞
u

∫0 e dx = 2 γ ⎜⎝ 2 , u ⎟⎠
− x2

⎛1 ⎞ • Clarifies relation between definition/integral

u u2 −x
γ ⎜ ,u2 ⎟ representation of probability integral/error
dx = ⎝ ⎠
2 1 e 2
∫ e dx = ∫
−x
function, Φ ( x ) & erf ( x ) (Formulas 8.250.1,
2

π 0 π 0 x π
8.251.1), and incomplete gamma function, which
= Φ (u ) = erf (u ) aids in derivation of formulas in Sect. 8.252,
[cf. 8.250 1, 8.251 1, 8.359 4] below, and GR.
• Both integrals solved by Formula 3.381.8,
m +1
∫ x e dx = (nβ ) γ (ν , βu ) ν = n
u
m − βx n ν −1 n

u > 0, Re ν > 0, Re n > 0, Re β > 0 ]

with parameters
⎡ 1 1⎤
⎢m = 0 or − 2 , n = β = 1 or 2,ν = 2 ⎥ and 8.359.4.
⎣ ⎦
π
( )
v v

∫e
−q2 x2
dx = [Φ(qv ) − Φ(qu )] Apply ∫x
m
exp − βx n dx , derived below, Sect. 3.326,
u
2q u

with parameters,

⎡ 1⎤
⎢⎣m = 0, n = 2, β = q ,ν = 2 ⎥⎦ ⇒
2

⎛1 ⎞ ⎛1 ⎞
Γ⎜ , q 2 u 2 ⎟ − Γ⎜ , q 2 v 2 ⎟
⎝2 ⎠ ⎝2 ⎠,
2q
⎛1 ⎞
and apply Formula 8.359. 3, Γ⎜ , x 2 ⎟.
⎝2 ⎠
May also be derived from indefinite integral

previously submitted, ∫ e − βx dx =
2 1 π
2 β
erf β x . [ ]
¾ Verified on Math.com
—Sect. 3.321——
 85

∞
π Apply previously derived Formula 3.381.9,
∫ e dx =
−q x
[1 − Φ(qu )] [q > 0]
2 2
∞

2q ∫x
m − βx n
e ( ) Γ(ν , βu )
dx = nβ ν
−1 n

u u

m +1
ν = [ Re n > 0, Re β > 0 ]
n
with parameters,
⎡ 1⎤
⎢⎣m = 0, n = 2, β = q , ν = 2 ⎥⎦ ⇒
2

1 ⎛1 2 2⎞
I= Γ⎜ , q u ⎟,
2q ⎝ 2 ⎠
⎛1 ⎞
and apply Formula 8.359 3, Γ⎜ , x 2 ⎟.
⎝2 ⎠
¾ Verified on Math.com
∞
π Since integrand is an even-function, set
∫e
−q 2 x 2
dx = ∞
I = 2 ∫ e − q x dx ,
2 2

−∞
q
0

& apply Formula 3.326.2 with parameters,

1
[m = 0, n = 2, β = q 2 , γ = ] .
2
• NOTE : useful for Gaussian probability density
+∞ 1 2
− x 1
function, ∫ e 2 dx = 2π , with q 2 = .
−∞
2
¾ Verified on Math.com
Formula 3.321 8 • Suggested new formula, 3.321.8
⎛ n +1 2 2 ⎞
γ⎜
,q u ⎟
• Derived from Formulas 3.381.8, 8.352.6, 8.356.3,
⎛n ⎞
u
⎝ 2 ⎠
∫0 x e dx = 2q n +1
n −q 2 x2
formula γ ⎜ , x ⎟ , derived in Sect. 8.359, below.
⎝2 ⎠
[u > 0, q > 0, n > −1] • ist formula is from Formula 3.381.8,
⎧ ( ) γ (ν , βu )
u
−1
∫x
m − βx n
e dx = nβ ν n

⎪
⎧ m ⎤⎫
0
⎪ ⎡ n −1

⎪ 1 ⎪⎛ n − 1 ⎞ ⎢ 2 (
q 2u 2 ⎥⎪) ν =
m +1
[ u > 0, Re ν > 0, Re n > 0, Re β > 0 ],
! 1− e −q u
∑ m! ⎥ ⎬
2 2

⎪ 2q n +1 ⎨⎜⎝ 2 ⎟⎠ ⎢ n
⎪ ⎪ ⎢
m=0
⎥⎦ ⎪⎭ ⎡ n + 1⎤
⎩ ⎣ with parameters ⎢m = n, n = 2, β = q 2 ,ν =
⎪
⎪ ⎣ 2 ⎥⎦
⎪⎡ n + 1 ⎤ • 2nd formula derived from new GR7 Formula
⎪⎢ int eger, n ≥ 1⎥
⎪⎣ 2 ⎦ 8.352.6, γ (n, x ) ,
⎪
=⎨ γ (n, x ) =
⎪
⎡1 ⎤ ⎡ ⎛ n−1 ⎞⎤
π Φ(qu ) −
m
⎪
⎛ ⎞ ⎢ ⎥ (n − 1)!⎢1 − e − x ⎜⎜ ∑ x ⎟ ⎥ [n = 1,2,...]
⎪ 1 ⎜ (n − 1)!!⎟
⎪ n +1 ⎜ n − 2 ⎟ ⎢
2
n−2
⎥ ⎣ ⎝ m =0 m! ⎟⎠ ⎦
⎪ 2 q ⎜
⎝ 2 2
⎠
⎢
⎟ − q 2u 2
⎢ qu ∑
2 (
2q u2 2
)s ⎥
⎥ n +1
⎪
⎢
e
( )( ) with n → integer, n = 1,3,5,..., x → q 2u 2 .
⎪ ⎣ s = 0 2 s + 1 2 s − 1 !!⎦⎥ 2
⎪ ⎡ n +1 ⎤ o Example, Formula 3.321.6,
⎪⎢ not int eger, n ≥ 0⎥
⎪⎣ 2 ⎦ u
n +1
∫0 x e dx, 2 = 2.
3 −q 2 x2
⎪
⎩
—Sect. 3.321——
 86

ƒ Use 8.352.6, γ (n, x ) , with

n +1
n→ , x → q 2u 2 , so that,
2
γ (2, q 2 u 2 )
=
2q 4
1 ⎡ 1
(
q 2u 2 )
m
⎤
⎢1 − e
2q 4 ⎣⎢
− q 2u 2
∑ m! ⎥
m =0 ⎦⎥
=
1
2q 4
[ (
1 − e −q u 1 + q 2u 2
2 2
)]
n +1
• 3rd formula for ≠ integer, n = 0,2,4,...
2
⎛ n +1 2 2 ⎞
γ⎜ ,q u ⎟
Solution for ⎝ 2 ⎠ in this case follows
n +1
2q
from proposed formula in Sect. 8.359, below,
⎛n ⎞
γ ⎜ , x⎟ =
⎝2 ⎠
⎡1
⎢2 π Φ x − ( ) ⎤
⎥
(n − 2)!! ⎢ n −3
⎥
⎢ ⎥
( )
n −3 1 2 s
2 2 ⎢e − x x 2 ∑ 2x ⎥
⎢⎣ s = 0 (2 s + 1)(2 s − 1)!!⎥
⎦
[n = 1,3,5,...]
by setting n → n + 1, x → q 2u 2 which leads to result:

⎛ n +1 2 2 ⎞
γ⎜ ,q u ⎟
⎝ 2 ⎠=
n +1
2q
⎡1 ⎤
π Φ(qu ) −
⎛ ⎞ ⎢2 ⎥
1 ⎜ (n − 1)!! ⎟ ⎢ n−2
⎥
⎜⎜ n − 2 ⎟⎟ ⎢
2q n +1
⎝ 2 2 ⎠ ⎢e − q u qu ∑
2 2
2 2 2 s
2q u ( ) ⎥
⎥
⎢⎣ s = 0 (2 s + 1)(2 s − 1)!!⎥
⎦
⎡n +1 ⎤
⎢ 2 not int eger, n ≥ 0⎥
⎣ ⎦
o Example:
n + 1 n=4 5
u

∫ x e dx,
4 −q x
= .
2 2
Formula 3.321.7,
0
2 2

—Sect. 3.321——
 87

⎛5 ⎞
γ ⎜ , q 2u 2 ⎟
⎝2 ⎠=
2q 5
⎡ ⎛1 ⎞⎤
⎢ ⎜ 2 π Φ(qu ) − ⎟⎥
1 ⎢3 ⎜ ⎟⎥
⎜
2 q ⎢ 2 − q 2u 2
5 1
( 2 2 s
2q u ) ⎟⎥
⎜
⎢ ⎜ e qu ∑ ⎟⎟⎥
s = 0 (2 s + 1)(2 s − 1)!! ⎠ ⎦
⎣ ⎝
⎡3 ⎤
⎢ π Φ(qu ) − ⎥
1 4
= 5⎢ ⎥
2 q ⎢ − q 2u 2 ⎛ 3 2 2 ⎞⎥
e qu⎜ + q u ⎟
⎢⎣ ⎝2 ⎠⎥⎦
⎛5 ⎞
NOTE: This agrees with calculation γ ⎜ , x ⎟ given
⎝2 ⎠
⎛n ⎞
below in Sect. 8.359, under γ ⎜ , x ⎟ .
⎝2 ⎠

o Example:
n + 1 n =6 7
u

∫ x e dx,
6 −q x
=
2 2

0
2 2
⎛7 ⎞
γ ⎜ , q 2u 2 ⎟
⎝2 ⎠=
2q 7
⎡15 ⎤
⎢ π Φ(qu ) − ⎥
1 8
⎢ ⎥
2q 7 ⎢ − q 2u 2 ⎛ 15 5 2 2 4 4 ⎞⎥
e qu⎜ + q u + q u ⎟
⎢⎣ ⎝4 2 ⎠⎥⎦
⎛7 ⎞
NOTE: This agrees with calculation γ ⎜ , x ⎟ given
⎝2 ⎠
⎛n ⎞
below in Sect. 8.359, under γ ⎜ , x ⎟ .
⎝2 ⎠

—Sect. 3.321——
 88

FORMULA PROOF OUTLINE

(formula references are to
G & R 7e)
v
⎛ x2 ⎞ • Complete the square, factoring the last term
∫u ⎜⎝ 4β ⎟⎟⎠dx =
exp⎜ − − γx
⎛ x ⎞
2

⎜
⎜2 β
+ γ β ⎟ − γ 2β ,
⎟
( )
⎡ ⎛ v ⎞⎟ ⎤ ⎝ ⎠
⎢Φ⎜⎜ γ β + ⎥
⎢ ⎝ 2 β ⎟⎠ ⎥ x
+γ β,
ds 1
(
πβ exp βγ ⎢)
2
⎥
s=
2 β
=
dx 2 β
,
⎢ ⎛ u ⎞⎥
⎜ ⎟
⎢− Φ⎜ γ β + 2 β ⎟⎥ use change of variable, su =
u
+γ β,
⎣ ⎝ ⎠⎦ 2 β
[Re β > 0] v
sv = +γ β,
2 β
to get,
sv

∫e
βγ 2
I = 2 βe −s 2
ds
su

• Apply Formula 2.33.10 for indefinite exponential

integral,
∫ x e dx, as shown below in Sect. 3.323 derivation
m − βx n

for
e − (ax + kbx + c )dx ⇒
2

∫
sv
βγ 2 ⎡ ⎛ 1 2 ⎞⎤ ,
u βe ⎢ − Γ⎜ 2 , s ⎟ ⎥
⎣ ⎝ ⎠ ⎦ su
⎛1 ⎞
• Apply Formula 8.359.3, Γ⎜ , x 2 ⎟ , to the
⎝2 ⎠
difference of
Φ ( su ) & Φ ( sv ).
¾ Verified on Math.com

—Sect. 3.322—
 89

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
∞

∫ exp(− ax )
Complete the square, factor the last term,
2
− bx dx = 2
b ⎞ ⎛ b2 ⎞ ⎛ b ⎞
2
⎛
0 a⎜ x + ⎟ − ⎜⎜ ⎟⎟ = ⎜⎜ a x + ⎟⎟
1 π ⎛ b 2 ⎞⎡ ⎛ b ⎞⎤ ⎝ 2a ⎠ ⎝ 4a ⎠ ⎝ 2 a⎠
exp⎜⎜ ⎟⎟ ⎢1 − Φ⎜⎜ ⎟⎟⎥ [a > 0]
2 a ⎝ 4a ⎠ ⎣ ⎝ 2 a ⎠⎦ ⎛ b2 ⎞
− ⎜⎜ ⎟⎟
⎝ 4a ⎠
use change of variable,
b ds b
s = ax + , = a , s0 = , to get
2 a dx 2 a
exp b4 a() 2
∞

∫e
−s2
I= ds =
a b
2 a

exp ( )Γ⎡⎢ 1 , ⎛⎜
b2
4a b ⎞ ⎤
2

⎟⎟ ⎥
⎜
2 a ⎣⎢ 2 ⎝ 2 a ⎠ ⎦⎥
by Formula 3.381.9,
∞

∫x
m − βx n
e (
dx = nβ ν ) Γ(ν , βu ) ,
−1 n

m +1
ν= [ Re n > 0, Re β > 0 ]
n
⎛1 ⎞
and apply Formula 8.359.3, Γ⎜ , x 2 ⎟ .
⎝2 ⎠
¾ Verified on Math.com
Complete the square, use change of variable,
∞

∫ exp(− ax )
b ds
2
− bx dx = s = ax + , = a,
u
2 a dx
⎛ b 2 ⎞⎡ b
1 π ⎛ b ⎞⎤ s u = au +
exp⎜⎜ ⎟⎟ ⎢1 − Φ⎜⎜ au + ⎟⎟⎥ 2 a
⎝
( )[(
2 a ⎝ 4a ⎠ ⎣ 2 a ⎠⎦
) ],
b2
[a > 0] to get , I =
exp
4a 1
Γ , au +
b 2
2 a 2 2 a

⎛1 ⎞
and apply Formula 8.359.3, Γ⎜ , x 2 ⎟ .
⎝2 ⎠
¾ Verified on Math.com

—Sect. 3.323—
 90

Similar to above formulas.

• Complete the square, use change of
∫ exp(− ax )
u
2
− bx dx = variable,
0
b ds
1 π ⎛ b 2 ⎞⎡ ⎛ b ⎞ ⎛ b ⎞⎤ s = ax + , = a,
exp⎜⎜ ⎟⎟ ⎢Φ⎜⎜ au + ⎟⎟ − Φ⎜⎜ ⎟⎟⎥ 2 a dx
2 a ⎝ 4a ⎠ ⎣ ⎝ 2 a⎠ ⎝ 2 a ⎠⎦ b b
s0 = , s u = au + ,
[u > 0, a > 0] 2 a 2 a
• Apply Formula 2.33.10 for indefinite
exponential integral,
∫ x e dx, as shown below in derivation for
m − βx n

⎛ b2 ⎞
exp⎜⎜ ⎟⎟ s
v
− (ax + kbx + c ) ⎝ 4a ⎠ ⎡ ⎛ 1 2 ⎞ ⎤ u
∫u e dx ⇒ ⎢ − Γ⎜ , s ⎟ ⎥
2

2 a ⎣ ⎝2 ⎠ ⎦ s0
⎛1 ⎞
• Apply Formula 8.359.3, Γ⎜ , x 2 ⎟ .
⎝2 ⎠

NOTE: above three integrals derivable from

following formula, with c = 0.
¾ Verified on Math.com

v
(ax ) • Complete the square, and factor last term
) ( ) (
2
+ kbx + c
∫e ( ) ( )
−
dx = kb 2
k 2b 2 −4 ac kb 2
k 2b 2 −4 ac
a x+ − = a x+ −
u 2a 4a 2 a 4a

⎡ ⎛ kb ⎞ ⎤ • Use change of variable on first term,

⎢ Φ ⎜
⎜ a v + ⎟⎟ ⎥ kb ds
1 π ⎛ k 2 b 2 − 4ac ⎞ ⎢ ⎝ 2 a⎠ ⎥ s = ax + ;
2 a dx
= a
exp⎜⎜ ⎟⎟
⎠⎢ kb ⎞⎥
kb kb
2 a ⎝ 4a ⎛ su = au + ;s =
2 a v
av +
⎢− Φ⎜⎜ au + ⎟⎟⎥
2 a

⎣⎢ ⎝ 2 a ⎠⎦⎥ • Then integral becomes,

k 2b 2 − 4 ac
[ a > 0] e 4a sv
∫ e − s ds
2
I =
a su

• Apply Formula 2.33.10 for indefinite

exponential integral,
(
Γ γ , βx n ) m +1
∫x e − βx dx = − γ =
n
m

nβ γ n

with x = s, m = 0, β = 1, n = 2, γ = 12 , and the

limits, s u & s v :
k 2b 2 − 4ac
I=
e 4a
2 a
−Γ 1 , s 2
2
[ ( )]ssuv
• Apply Formula 8.359.3, Γ 1 , x 2 ,
2
( )
to the difference of Φ ( su ) & Φ ( sv ).
I =

1
2 a
(
exp k b −4 ac
2 2
4a
){ ππ [[11−−ΦΦ((ss ))]]−} u
v
—Sect. 3.323—
 91

• Simply and integral results

• Since Φ (∞ ) =1, the integrals of this
section are obtained by substituting
u, v, su , sv & k .
¾ Verified on Math.com

NOTE: formula derivable from indefinite

integral, Formula 2.33.1, with k set to 2.
(ax )
( )
v 2
+ kbx + c
− ax + 2 bx + c
2 e− dx =
∫e dx = Set k = 2 in v ⎡ ⎛
⎢Φ ⎜ a v +
kb ⎞ ⎤
⎟ −⎥
u
∫1 π ⎛ k 2b 2 − 4ac ⎞ ⎢ ⎝
exp⎜⎜ ⎟⎟
2 a⎠ ⎥
⎡ ⎛ b ⎞ ⎤
u
2 a ⎝ 4a ⎠ ⎢Φ⎛ au + kb ⎞ ⎥
⎢ Φ ⎜
⎜ a v + ⎟⎟ ⎥ ⎢ ⎜ ⎟ ⎥
2 a ⎠ ⎦⎥
1 π ⎛ b 2 − ac ⎞ ⎢ ⎝ a⎠ ⎥ ⎣⎢ ⎝
exp⎜⎜ ⎟⎟
2 a ⎝ a ⎠⎢ ⎛ b ⎞⎥
⎢− Φ⎜⎜ au + ⎟⎟⎥
⎣⎢ ⎝ a ⎠⎦⎥
[a > 0]

∞
( ) Follows from above integral ( )
v
− ax + kbx + c
2

− ax + kbx + c ∫e
2

∫e dx = dx
u

⎛ kb ⎞
0

1 π ⎛ k 2 b 2 − 4ac ⎞ ⎡ ⎛ kb ⎞⎤ u = 0, v = ∞, Φ (u ) = Φ⎜⎜ ⎟⎟,

exp⎜⎜ ⎟⎟ ⎢1 − Φ⎜⎜ ⎟⎟⎥ with ⎝2 a ⎠
2 a ⎝ 4 a ⎠⎣ ⎝ 2 a ⎠⎦ Φ (v ) = Φ (∞ ) = 1.
[a > 0] ¾ Verified on Math.com
∞
(ax 2
+ bx + c ) Follows from above integral v e − (ax + kbx+c )dx
∫u
2

∫e
−
dx =
⎛ b ⎞
0

1 π ⎛ b 2 − 4ac ⎞ ⎡ ⎛ b ⎞⎤ k = 1, u = 0, v = ∞, Φ (u ) = Φ⎜⎜ ⎟⎟,

exp⎜⎜ ⎟⎟ ⎢1 − Φ⎜⎜ ⎟⎟⎥ with ⎝2 a ⎠
2 a ⎝ 4a ⎠ ⎣ ⎝ 2 a ⎠⎦ Φ (v ) = Φ (∞ ) = 1.
[a > 0] ¾ Verified on Math.com
∞
(
− ax + bx + c
2
) Follows from above integral v e − (ax + kbx+c )dx
∫u
2

∫e dx =
u
k = 1, u = u, v = ∞, Φ(u ) = Φ (u ),
1 π ⎛ b 2 − 4ac ⎞ ⎡ ⎛ b ⎞⎤ with
exp⎜⎜ ⎟⎟ ⎢1 − Φ⎜⎜ au + ⎟⎟⎥ Φ (v ) = Φ (∞ ) = 1.
2 a ⎝ 4a ⎠ ⎣ ⎝ 2 a ⎠⎦ ¾ Verified on Math.com
[a > 0]

—Sect. 3.323—
 92

u
( ) Follows from above integral ( )
v
− ax + kbx + c
2

− ax + kbx + c ∫e
2

∫e dx = dx
u
0
with
⎡ ⎛ kb ⎞ ⎤ ⎛ kb ⎞
⎢Φ⎜⎜ au + ⎟⎟ − ⎥ u = 0, v = u , Φ(u ) = Φ⎜⎜ ⎟⎟,
1 π ⎛ k b − 4ac ⎞ ⎢ ⎝ 2 a⎠ ⎥ 2 2
⎝2 a ⎠
exp⎜⎜ ⎟⎟
2 a ⎝ 4a ⎠ ⎢ ⎛ kb ⎞ ⎥ ⎛ kb ⎞
⎢Φ⎜⎜ ⎟⎟ ⎥ Φ (v ) = Φ⎜⎜ au + ⎟⎟.
⎣⎢ ⎝ 2 a ⎠ ⎥⎦ ⎝ 2 a⎠
[u > 0, a > 0] ¾ Verified on Math.com

∞
(ax )
∫ (cx )
2
+ jbx A long calculation similar to other forms in
2
+ kgx e − dx = Sect. 3.323.
−∞
¾ math.com calculation
π ⎛ j 2b 2 ⎞
[(
a −5 2 c b 2 j 2 + 2a − 2abgjk exp⎜⎜ ) ] ⎟⎟
4 ⎝ 4a ⎠
[a > 0]
∞
(ax )
∫ (cx )
2
+ bx A long calculation similar to other forms in
2
+ gx e − dx = Sect. 3.323.
0
¾ math.com calculation
⎛ b 2 ⎞⎡ ⎛ b ⎞⎤
1
[
π a −5 2 b 2 c + 2a(c − bg ) exp⎜⎜ ⎟⎟ ⎢1 − Φ⎜⎜ ⎟⎟⎥ ]
8 ⎝ 4a ⎠ ⎣ ⎝ 2 a ⎠⎦
2ag − bc
+
4a 2

[a > 0]
∞
( px )
∫ (ax )
2
+ kqx Adapted from—A.P. Prudnikov, et al.,
2
+ bx + c e − dx = Integrals and Series, Vol.1: Elementary
−∞
Functions, 1986, page 344, Formula
π ⎡4cp 2 − 2kbpq ⎤ ⎛ k 2q2 ⎞ 2.3.15.12
⎥ exp⎜⎜ ⎟⎟
−5 2
p ⎢
4 ⎣⎢+ a k q + 2 p ⎦⎥
2 2
⎝ 4p ( ) ⎠
¾ verified on math.com
[Re p > 0]
∞
( px )
∫ (ax )
2
+ qx + c Adapted from—A.P. Prudnikov, et al.,
2
+ kbx e − dx = Integrals and Series, Vol.1: Elementary
−∞
Functions, 1986, page 344, Formula
π ⎛ q 2 − 4 pc ⎞
[(
p −5 2 a q 2 + 2 p − 2kbpq exp⎜⎜ ⎟⎟ ) ] 2.3.15.12
¾ verified on math.com
4 ⎝ 4p ⎠

[Re p > 0]
∞
( )
∫ (ax + kbx+ c)e
− px + jqx+r
2 Adapted from—A.P. Prudnikov, et al.,
2
dx = Integrals and Series, Vol.1: Elementary
−∞
Functions, 1986, page 344, Formula
π ⎡4cp2 − 2 jkbpq +⎤ ⎛ j 2 q2 − 4 pr ⎞ 2.3.15.12
−5 2
⎢ ⎥ ⎜⎜ ⎟⎟
( )
p exp ¾ verified on math.com
⎣⎢ a j q + 2 p ⎦⎥
2 2
4 ⎝ 4p ⎠

[Re p > 0]
—Sect. 3.323—
 93

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
⎛ a+b⎞ ¾ Set b to a + b in ∫ exp⎛⎜ − b2 ⎞⎟dx ,
u u

∫0 exp⎜⎝ − x 2 ⎟⎠dx = 0 ⎝ x ⎠

⎡ Sect. 3.471, below

⎛ a+b⎞ ⎛ ⎞⎤
u exp⎜ − 2 ⎟ − (a + b )π ⎢1 − Φ⎜⎜ a + b ⎟⎟⎥ ¾ Verified on Math.com
⎝ u ⎠ ⎢⎣ ⎝ u ⎠⎥⎦
[u > 0, Re(a + b ) > 0]

—Sect. 3.324—
 94

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
v
( Γ γ , βu n − Γ γ , βv n
) ( ) ( ) •
v

∫x (
exp − βx n dx ) is a generalization of
∫u x exp − βx dx =
m n m

nβ γ u

contributor’s Formula 3.326.2. Follows from

m +1
γ = , [Re β > 0, Re n > 0] indefinite integral,2.33.10,
n (
Γ γ , βx n ) m +1
∫ x e dx = − nβ γ
m − βx n
γ = ,
n
by taking limits u and v.
v
(
exp − β x n ) (
Γ z, βu n − Γ z , βv n ) ( )
∫u x m dx =
nβ z
NOTE: All previously submitted formulas for
this integrand are derivable by substituting the
1− m quantities u , v, su , sv and the properties of the
z= , [Re β > 0, Re n > 0]
n incomplete gamma function for real variables,
Γ(a, x ) :
Γ(a,0) = Γ(a); Γ(a, ∞ ) = 0,
Γ(a ) = Γ(a, x ) + γ (a, x ).
•
v
exp − βx n ( )
∫u x m dx is from integral in Sect. 3.35,
below,
⎛ a ⎞ ⎛ a ⎞
n+b Γ⎜ z m , β u n + b ⎟ − Γ⎜ z m , βv n + b ⎟
v
e − βx ⎝ n+b ⎠ ⎝ n+b ⎠
∫u x m ± a dx = a

(n + b )β
zm
n+b

1− m
z= , [Re β > 0, Re n + b > 0]
n+b
with parameters [a = b = 0] .
¾ Verified on math.com
Formula 3.326.2 Apply above formula,
∞
Γ (γ ) ( ) ( ) (
Γ γ , βu n − Γ γ , βv n )
( )
v

∫0 x exp − β x dx = n β γ ∫ x exp − βx dx =
m n
m n
u
nβ γ
m +1
m +1 γ=
γ = , [Re n > 0 , Re β > 0 , Re γ > 0 ] n
n Noting Γ(a,0 ) = Γ(a ); Γ(a, ∞ ) = 0

v
Factor e −b and apply above integral
∫x
m
exp(− ρx − b )dx = v
Γ(γ , βu n ) − Γ(γ , β v n )
γ=
m +1
[Re β > 0, Re n > 0]
∫x e − βx dx =
n
m

u nβ γ n

with parameters [m = m, n = 1, β = ρ , γ = m + 1]
u

Γ(m + 1, ρu ) − Γ(m + 1, ρv )
e −b [Re ρ > 0] ¾ Verified on math.com
ρ m +1

—Sect. 3.326—
 95

∞
( )
Use Formula 3.326.2 with parameters,
∫ x exp − βx dx =
m + n +1 n +1

⎡m = m + n + 1, n = n + 1, β = β ,⎤
0

(m + 1)Γ⎛⎜ m + 1 ⎞⎟ ⎢ ⎥
⎝ n +1 ⎠ ⎡ m +1 ⎤ ⎢γ = m + n + 1 + 1 ⎥
m +1 ⎢Re β > 0, Re n + 1 > 0⎥ ⎣ n +1 ⎦
+1 ⎣ ⎦
(n + 1)2 β n +1
⎛ m +1 ⎞
Γ⎜ + 1⎟
⎝ n +1 ⎠
to obtain , leading to integral.
m +1
+1
(n + 1) β n+1
Verified on math.com
• Separate variables (x − a ) , use change of
Formula 3.326.3 variable, s = x − b, ds = dx
∞
s 0 = −b, to get
∫ ( x − a )e
− β ( x −b ) n
dx =
∞ ∞

∫ se ds − (a − b ) ∫ e ds ,
0
− βs n
− βs n
and
⎛2 n⎞ ⎛1 n ⎞
Γ⎜ , β (− b ) ⎟ Γ⎜ , β (− b ) ⎟ −b −b
⎝n ⎠ − (a − b ) ⎝ n ⎠ apply twice derived integral above,
2 1
∞
m +1
nβ n nβ n
∫u x e dx = (nβ ) Γ(ν , βu ) ν = n
m − βx n ν −1 n

[ Re n > 0, Re β > 0, arg b < π ]

¾ May also be solved by repeated
application of previous formula for
indefinite exponential integrals:
Γ γ , βx n ( m +1 )
∫
m − βx n
x e dx = − γ
, γ =
nβ n
Formula 3.326.4 Similar to above integral
u

∫ ( x − a )e
− β ( x −b ) n
dx = ∞

∫ ( x − a )e
− β ( x −b ) n
0 dx
⎛2 n ⎞ ⎛2 n ⎞
Γ⎜ , β (− b ) ⎟ − Γ⎜ , β (u − b ) ⎟
0

⎝n ⎠ ⎝n ⎠−
2

nβ n
⎛1 n ⎞ ⎛1 n ⎞
Γ⎜ , β (− b ) ⎟ − Γ⎜ , β (u − b ) ⎟
(a − b ) ⎝ n ⎠ ⎝n ⎠
1

nβ n
[ u > 0, Re n > 0, Re β > 0, arg b < π , arg u-b < π ]

—Sect. 3.326—
 96

Formula 3.326.5 Similar to above integral

∞

∫ ( x − a )e
− β ( x −b )
dx =
n
∞

∫ ( x − a )e
− β ( x −b ) n
u dx
⎛2 n ⎞ ⎛1 n ⎞
Γ⎜ , β (u − b ) ⎟ Γ⎜ , β (u − b ) ⎟
0

⎝n ⎠ − (a − b ) ⎝ n ⎠
2 1

nβ n nβ n
[ Re n > 0, Re β > 0, arg u − b < π ]

∞ Similar to above integral

∫ ( x + a )e
− β ( x −b )
dx =
n

0 ∞

∫ ( x − a )e
− β ( x −b ) n
⎛2 n ⎞ ⎛1 n⎞ dx
Γ⎜ , β (− b ) ⎟ Γ⎜ , β (− b ) ⎟
⎝n ⎠ + (a + b ) ⎝ n ⎠
0

2 1

nβ n
nβ n

[ Re n > 0, Re β > 0, arg b < π ]

∞ Similar to above integral

∫ ( x + a )e
− β ( x −b )
dx =
n

u ∞

∫ ( x − a )e
− β ( x −b ) n
⎛2 n⎞ ⎛1 n⎞ dx
Γ⎜ , β (u − b ) ⎟ Γ⎜ , β (u − b ) ⎟
⎝n ⎠ + (a + b ) ⎝ n ⎠ 0

2 1

nβ n
nβ n

[ Re n > 0, Re β > 0, arg u − b < π ]

u
Similar to above integral
∫ ( x + a )e
− β ( x −b )
dx =
n

0 ∞

∫ ( x − a )e
− β ( x −b ) n
⎛2 n ⎞ ⎛2 n ⎞
Γ⎜ , β (− b ) ⎟ − Γ⎜ , β (u − b ) ⎟
dx
⎝n ⎠ ⎝n ⎠+ 0
2

nβ n
⎛1 n ⎞ ⎛1 n ⎞
Γ⎜ , β (− b ) ⎟ − Γ⎜ , β (u − b ) ⎟
(a + b ) ⎝ n ⎠ ⎝n ⎠
1

nβ n
[ u > 0, Re n > 0, Re β > 0, arg b < π , arg u-b < π ]

—Sect. 3.326—
97
 98

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
v
( ) ( ) (
Ei − ae nv − Ei − ae nu ) • Change of Variable:
∫ exp − ae dx =
nx

n s = e nx , ds = ne nx dx = nsdx, su = e nu , s v = e nv
u

[n > 1, Rea > 0] 1 v e − as

I= ∫
s
1 sv
ds = Ei(− as ) =
n su s n su
1
n
[ ( ) (
Ei − ae nv − Ei − ae nu )]
by 2.33.18.
NOTE: if
Ei(− ∞ ) − Ei(− a )
= − Ei(− a )
1
u = 0, v = ∞, I =
n n
, Formula 3.327.
NOTE: present formula 3.327 constraints
need changing by deleting a ≠ 0 to remove
redundancy.

¾ Math.com unable to do but returned

derived indef. integral,

( ) (
Ei − ae nx )
∫ exp − ae dx = n
nx

∞ Derived Sect. 8.367, below.

− (x + e − x )
∫
0
xe dx = C − Ei(−1)

∞ Derived Sect. 8.367, below.

− (x − e − x )
∫ xe
0
dx = Ei(1) − C

—Sect. 3.327—
 99

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
• Re-express integral,
∫ exp(− e )e ( ) ( )
v
μ
x x
dx = Γ μ , e u − Γ μ , e v
( )
v
μ
I = ∫ e −e e x dx
x
u

• Change of variable:
ds
s = ex , = e x = s , su = e u , s v = e v
dx
sv − s μ sv
e s
I=∫ ds = ∫ s μ −1e − s ds =
su
s su

s = sv
− Γ (μ , s )
s = su
( ) (
= Γ μ , eu − Γ μ , ev )
by indefinite integral 2.33.10,

(
Γ γ , βx n ) m +1
∫x
m − βx n
e dx = − γ = ,
nβ γ n
with parameters [m = μ − 1, β = n = 1, γ = μ ]
NOTE: if
u = −∞, v = +∞, I = Γ(μ ,0 ) − Γ(μ , ∞ ) = Γ(μ )
, Formula 3.328

¾ Not verifiable on math.com, but

integral with limits, u = 0, v = ∞
gives I = Γ(μ ,1).

∫ exp(− e )e ( ) ( )
v
μ
Similar to above integral,
−x x
dx = Γ − μ , e −v − Γ − μ , e −u
∫ exp(− e )e ( ) ( )
v
μ
u
x x
dx = Γ μ , e u − Γ μ , e v .
u

Re-express integral as,

( )
v
−μ
∫e e
−x
e −x
dx , and make change of
u

variable, s = e − x .

—Sect. 3.328—
 100

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
∞ ¾ Euler’s Constant
∫ x exp(x − e )dx = −C
x
¾ Derived in Sect. 8.367, below.
−∞

( )
v Similar to derivation of Sect. 3.328,
∫ exp − βe − μx dx =
−x

∫ exp(− e )e ( ) ( )
v
μ
u
x x
dx = Γ μ , eu − Γ γ , e v by re-
β −μ
[Γ(μ , βe ) − Γ(μ , βe )]
−v −u u

[Re β > 0] expressing integral as,

∫e
− βe− x
( ) μ
e − μx dx = ∫ e − βe e − x dx , and making
−x

change of variable, s = e − x .
If u = 0, v = ∞,

I=
( ) (
Γ μ , βe −∞ − Γ μ , βe −0 )=
μ
β
NOTE:
Γ(μ ,0) − Γ(μ , β ) γ (μ , β )
= ,
βμ βμ
Formula 3.331 1, by Formula 8.356 3.
¾ Math.com could not solve, but did solve case
of u = 0, v = ∞.

∫ exp(− βe )
v Similar to above integral,
x
− μx dx =
∫ exp(− βe )
v
u
−x
− μx dx =
β [Γ(− μ , βe ) − Γ(− μ , βe
μ u v
)] u

[Re β > 0] β − μ [Γ(μ , βe −v ) − Γ(μ , βe −u )]

by re-expressing integral as, ∫ e − βe e x
x
( ) −μ
dx and
making change of variable, s = e x .

NOTE: if u = 0, v = ∞, I = β μ Γ(− μ , β ),
Formula 3.331 2
¾ Math.com could not solve, but did solve
case of u = 0, v = ∞.

∫ exp(− βe )
v Similar to above integrals, by re-expressing
+ μx dx =
( )
x
μ
integral as, ∫ e − βe e x
x

u
dx and making change of
β − μ [Γ(μ , βe u ) − Γ(μ , βe v )] variable, s = e x .
[Re β > 0]
∞

∫ exp(− βe )
In above integral,
x
+ μx dx = β − μ Γ(μ , β )
∫ exp(− βe )
v
0
x
+ μx dx =
[Re β > 0] u

β − μ [Γ(μ , βe u ) − Γ(μ , βe v )],

set u = 0, v = ∞.
—Sect. 3.331—
 101

NOTEs:
• A useful integral transform for deriving
dΓ(a, x )
by Leibnitz’s rule. See Sect. 8.356
da
• Form. 3.331.1 is likewise a useful integral
dγ (a, x )
transform for deriving by Leibnitz’s
da
rule. See Sect. 8.356

∫ exp(− βe )
v Similar to above integrals by re-expressing
−x
+ μx dx =
u
integral as, ∫ e − βe
−x
(e )
−x −μ
dx and making change
β μ [Γ(− μ , βe −v ) − Γ(− μ , β e −u )] of variable, s = e − x .
[Re β > 0]
∞ Exercises
Γ(μ , β ) = β μ ∫ x μ −1e − βx dx • Check solutions on math.com
1
NOTE: Useful transforms for derivatives below.
1
γ (μ , β ) = β μ ∫ x μ −1e − βx dx
0
∞
Γ(μ ) = β μ ∫ x μ −1e − βx dx (GR 8.312.2)
0

See Sect. 8.356, below.

γ (μ , β )
d
dβ

Γ(μ , β )
d
dβ

—Sect. 3.331—
102
 103

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
Formula 3.351.4 Derived from indefinite integral,2.33.10,
Γ(z , βu n )
∞ − βx n
v − ρx
e e 1− m
∫u x n+1 dx ∫u x m dx =
nβ z
z=
n
[Re β > 0, Re n > 0]

Add Formula with parameters [β = ρ , m = n + 1, n = 1, z = −n]

= ρ n Γ(− n, ρu )
Formula 3.351.6 From definition & properties of exponential-integral
Add Formula function,
∞ −x
x=∞
I = −Ei(− x ) = −[0 − Ei(u )] = Ei(u )
e
= −∫ dx
−u
x x = −u
Formula 3.351.7 Derived from
m +1
( )
u u

∫x dx = (nβ ) γ ν , βu n ,ν =
−1
∫ xe
− μx m − βx n
dx e
0 0
n
Add Formula [u > 0, Reν > 0, Re n > 0, Re β > 0]
= μ −2 Γ(2, μu ) with parameters [β = μ , n = 1, m = 1,ν = 2]

Formula 3.351.8 u

∫ xe
− μx
u Similar to Formula 3.351.7, dx
∫ x e dx
2 − μx
0
0

Add Formula
= μ −3 Γ(3, μu )
Formula 3.351.9 u

∫ xe
− μx
u Similar to Formula 3.351.7, dx
∫ x e dx
3 − μx
0
0

Add Formula
= μ −4 Γ(4, μu )
Γ(n + 1,− μu ) − Γ(n + 1,− μv ) •
v
Generalizes above forms
∫ x e dx =
n − μx

μ n+1 • Derived from formula in Sect. 3.326, above

u v
( ) ( ) (
Γ γ , βu n − Γ γ , βv n ) m +1
∫ x exp − βx dx = γ=
m n

u
nβ γ n
[Re β > 0, Re n > 0] with parameters
[m = n, β = μ , n = 1, γ = n + 1]

—Sect. 3.351—
 104

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
v
xe x ev eu Use change of variable,
∫u (1 + x )2 dx = v + 1 − u + 1 s = x + 1; ds = dx; su = u + 1; sv = v + 1; x = s − 1,
to give,
⎡ su e s su s
e ⎤
I = e −1 ⎢ ∫ ds − ∫ 2 ds ⎥ .
⎣⎢ sv s sv
s ⎦⎥
Do integration by parts on first integral, to get,
su s s
e e s sv u e s
∫ s ds = s su + s∫ s 2 ds , so that
sv v

⎡ e s s v su e s su s
e ⎤
I = e −1 ⎢ + ∫ 2 ds − ∫ 2 ds ⎥ =
⎢⎣ s s u sv s sv s ⎥⎦
−1 ⎡ e e u +1 ⎤
v +1
ev eu
e ⎢ − ⎥= −
⎣ v + 1 u + 1⎦ v + 1 u + 1
NOTEs:
v
xe x
• If u = 0, v = 1 in ∫ dx, integral 3.353 4 results
u (1 + x )
2

• Neither of the following two integrals give such simple

solutions:
xe − x
v v
xe x
o ∫u (1 + x )2 dx & ∫u (1 − x )2 dx
¾ Verified on math.com

xe − x e −v e −u
v v
xe x ev eu
∫u (1 − x )2 dx = −
1− v 1− u
Approach and proof similar to ∫u (1 + x )2 dx = −
v +1 u +1

¾ Verified on math.com

—Sect. 3.353—
105
 106

FORMULA PROOF OUTLINE

(formula references are to G & R
7e)

[Φ( qv )− Φ( qu )]
v − βx n +b
v
e − qx π
∫u x dx = [Re q > 0] e
Derived from ∫ m ± a dx , Sect.
q u
x
3.462, below, with parameters,
⎡ 1 ⎤
⎢ β = q, n + b = 1, m = 2 ,± a = 0,⎥
⎢ ⎥
⎢z = 1 − m = 1 ⎥
⎢⎣ 2 ⎥⎦
giving,
⎛1 ⎞ ⎛1 ⎞
Γ⎜ , qu ⎟ − Γ⎜ , qv ⎟
e − qx
v
⎝2 ⎠ ⎝2 ⎠=
∫u x dx = 1

q2
π
q
{[1 − Φ( qu )]− [1 − Φ( qv )]} ,

applying newly suggested formula

⎛1 ⎞
in Sect. 8.359, below, Γ⎜ , x ⎟ .
⎝2 ⎠
¾ Verified on math.com
Set v → ∞, Φ (v ) = 1 in above
[1 − Φ( qu )]
∞
e − qx π
∫u x dx = [Re q > 0] integral,
q
e − qx π
[Φ( qv ) − Φ( qu )]
v

∫
u x
dx =
q

—Sect. 3.361—
 107

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
π ⎡ ⎛⎜ ρb ⎞⎟⎤
ρb
v
e − ρx ± ρb ⎞⎟ ⎛ Below in Sect. 3.3462 formula,
∫ ax ± b
dx = e a
⎢ Φ ⎜ ρv ±
aρ ⎢⎣ ⎝ a ⎟⎠
− Φ ⎜ ρu ±
⎜ ⎥
a ⎟⎠⎥⎦ ±
ρb ⎡ ⎛ ρ⎞ ⎤
⎢Γ⎜ − n + 1, ρu ± ba ⎟ −⎥
⎝ v
e − ρx ρ n −1e a
⎢ ⎝ ⎠ ⎥
∫u (ax ± b)n dx = a n
u
,
⎢ ⎛ ρ⎞ ⎥
⎡ ⎛b ⎞ ⎛b ⎞ ⎤ Γ
⎢ ⎜ − n + 1, ρv ± ⎟ ⎥
⎢ ρ > 0, arg⎜ ± u ⎟ < π , arg⎜ ± v ⎟ < π ⎥ ⎣ ⎝ ba ⎠ ⎦

⎣ ⎝a ⎠ ⎝a ⎠ ⎦
1
set parameter, n = ,
2
⎛1 ⎞
and apply, Γ⎜ , x ⎟, Sect. 8.359, below.
⎝2 ⎠
EXAMPLE: Formula 3.362 2,
∞
e − μx
∫0 x + β dx ,
set parameters,
[ρ = μ , a = 1, b = β , n = 1 / 2, su = 0, sv = ∞]
and solve, recognizing Φ (∞ ) = 1.
¾ Verified on math.com
Set v → ∞, ρ = μ , a = 1, b = β in
[ ( )]
∞
e − μx π βμ
∫
u x+β
dx =
μ
e 1 − Φ μ u + βμ above integral,
⎡ ⎛ ρb ⎞⎟ ⎤
[Re μ > 0, arg(u + β ) < π ] v
e− ρx ±
ρb
⎢Φ⎜⎜ ρv ±
π ⎢ ⎝
−⎥
a ⎟⎠ ⎥
∫ ax ± b
dx = e a
aρ ⎢⎢ ⎛ ρb ⎞⎟ ⎥
⎥
u
⎜
Φ ρu ±
⎢ ⎜ a ⎟⎠ ⎥⎦
⎣ ⎝
¾ Verified on math.com
Set u = 0, v = u , ρ = μ , a = 1, b = β
∫ x + β dx = μ e [Φ( β u + βμ ) − Φ( βμ )]
− μx
u
e π βμ
in above integral,
0
⎡ ⎛ ρb ⎞⎟ ⎤
[u > 0, Re μ > 0, arg(u + β ) < π , arg(β ) < π ] v
e− ρx ±
ρb
⎢Φ⎜⎜ ρv ±
π ⎢ ⎝
−⎥
a ⎟⎠ ⎥
∫ ax ± b
dx = e a
aρ ⎢⎢ ⎛ ρb ⎞⎟ ⎥
⎥
Φ⎜⎜ ρu ±
u
⎢
⎣ ⎝ a ⎟⎠ ⎥⎦
¾ Verified on math.com
∞ ρb
− μx
π −μu e − ρx π
v
[Re μ > 0]
e ±
∫
u x−u
dx =
μ
e From ∫
u ax ± b
dx = e a
aρ
with u = u, v = ∞ substituted

—Sect. 3.362—
 108

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
∞
π (2μu + 3) Change of variable; substitute
∫x x − u e − μx dx = 5
e −uμ [Re μ > 0] s 2 = x − u; x = s 2 + u
u
4μ 2
dx = 2sds; su = 0; s∞ = ∞
to obtain,
∞
( )
I = 2e −uμ ∫ s 2 s 2 + u e − μs ds =
2

⎡ ⎛5⎞ ⎛ 3 ⎞⎤
⎢ Γ⎜ 2 ⎟ Γ⎜ ⎟ ⎥
2e −uμ ⎢ ⎝ 5⎠ + u ⎝ 3⎠ ⎥
2
⎢ ⎥
⎢ 2μ 2μ 2 ⎥
2

⎣ ⎦
by use of Formula 3.326.2 applied twice,
Γ(γ , β u n )
∞

∫0 x m
exp (− β x n
)dx =
nβ γ
m +1
γ = , [Re n > 0, Re β > 0, Re γ > 0]
n
with parameters
⎡ 3 5⎤
⎢⎣m = 2 & 4, n = 2, β = μ , γ = 2 & 2 ⎥⎦ ,
⎛5⎞ 3 ⎛3⎞ 1
noting that Γ⎜ ⎟ = π & Γ⎜ ⎟ = π .
⎝2⎠ 4 ⎝2⎠ 2
Then, combing and simplifying results,
∞
π (2 μu + 3) −uμ
∫ x x − u e − μx dx = 5
e [Re μ > 0]
u
4μ 2
NOTE: other integrals in Sect. 3.363 are
more challenging.
¾ Verified on math.com
v Generalization of above integral,
∫x a + bxe − px dx = ∞
π (2μu + 3) −uμ
∫x
− μx
u x − ue dx = 5
e
⎧⎡ ⎛ 5 p ⎞ ⎛5 p ⎞⎤ ⎫ u
4μ 2

⎪⎢Γ⎜ , (a + bu )⎟ − Γ⎜ , (a + bv )⎟⎥ − ⎪ [Re μ > 0]

pa
e b
⎪⎣ ⎝ 2 b ⎠ ⎝2 b ⎠⎦ ⎪
⎨ ⎬
pa ⎡ 3 p ⎤ • Use change of2 variable:
5
p 2 b ⎪ ⎢Γ⎛⎜ , (a + bu )⎞⎟ − Γ⎛⎜ , (a + bv )⎞⎟⎥ ⎪
3 p
⎪ b ⎣ ⎝2 b s −a
⎠⎦ ⎪⎭ s 2 = a + bx; x =
s
⎩ ⎠ ⎝2 b ; dx = 2 ds
b b
su = a + bu ; sv = a + bv
[Re p > 0, b > 0] to give integral,

—Sect. 3.363—
 109

pa
⎛ s 2 − a ⎞ − pba
sv
2e b
I= ∫ s ⎜⎜⎝ b ⎟⎟⎠e ds
2

b su
pa
⎡ sv 4 − ps ⎤
s s
ps
2e b sv −
= 2 ⎢ ∫ s e ds − a ∫ s e b ds ⎥
b 2

b ⎢⎣ su su
⎥⎦
pa

2e b ⎡ sv 4 − pb −
pb ⎤
= 2 ⎢ ∫ s e a − as e a ds ⎥
2

b ⎢⎣ su ⎥⎦
⎡ ⎛5 p 2⎞ ⎛5 p 2⎞ ⎤
⎢ Γ⎜ 2 , b s u ⎟ − Γ⎜ 2 , b s v ⎟ ⎥
⎢ ⎝ ⎠ ⎝ ⎠ −⎥
⎢ 5
⎥
pa ⎢ ⎛ p ⎞2 ⎥
2⎜ ⎟
2e b ⎢ ⎝b⎠ ⎥
= 2 ⎢ ⎥
b ⎢ ⎛3 p 2⎞ ⎛3 p ⎞
Γ⎜ , s u ⎟ − Γ⎜ , s 2v ⎟ ⎥
⎢ ⎝2 b ⎠ ⎝2 b ⎠ ⎥
⎢ 3 ⎥
⎢ ⎛ ⎞
p 2 ⎥
⎢ 2⎜ ⎟ ⎥
⎣ ⎝b⎠ ⎦
by integral, Sect. 3.326, above,
v
( ) ( ) (
Γ γ , βu n − Γ γ , βv n )
∫u − β =
m n
x exp x dx
nβ γ
m +1
γ = , [Re β > 0, Re n > 0]
n

with parameters
[m = 4 & 2, n = 2, β = p b , γ = 5 2 & 3 2]
Simply for final result. May be simplified
further by using results of Section 8.359, viz.,

⎛3 ⎞ 1 ⎛1 ⎞ − x 2
1
Γ ⎜ , x ⎟ = Γ⎜ , x ⎟ + e x
⎝2 ⎠ 2 ⎝2 ⎠
⎛5 ⎞ 3 ⎛1 ⎞ ⎛ ⎞ 1
Γ⎜ , x ⎟ = Γ⎜ , x ⎟ + e − x x 2 ⎜ + x ⎟.
⎜3 ⎟

⎝2 ⎠ 4 ⎝2 ⎠ ⎜2
⎝
⎟
⎠

NOTE: Apply integral to solution of

∞
π (2 μu + 3) −uμ
∫u x x − u e dx =
− μx
5
e
4μ 2

[Re μ > 0]
to verify solution.

—Sect. 3.363—
 110

v
A further generalization of above integral,
∫ x(a + bx ) e dx =
− px
m
∞
π (2μu + 3) −uμ
∫u x x − u e dx =
− μx
u
5
e
⎧⎡ ⎛ ⎞ ⎤ ⎫ 4μ 2
⎪⎢Γ⎜ m + 2, (a + bu )⎟ − ⎥ ⎪
p
⎪⎢ ⎝ b ⎠ ⎥ ⎪
−
[Re μ > 0]
⎪⎢ ⎛ ⎞ ⎥ ⎪ Left as exercise.
⎪⎢Γ⎜ m + 2, (a + bv )⎟ ⎥ ⎪
pa
p
∞
e b
⎪⎣ ⎝ b ⎠ ⎦ ⎪
m + 2 1− m ⎨ ⎬ Check against ∫x x − u e − μx dx with
p b ⎪ ⎡Γ⎛ m + 1, p (a + bu )⎞ − ⎤ ⎪ u
⎪ pa ⎢ ⎜⎝ b
⎟ ⎥⎪
⎠ ⎥ 1
⎪ ⎢ ⎪ m= , noting properties of complete gamma
⎪b ⎢ ⎛ ⎞ ⎥⎪ 2
⎪ ⎢Γ⎜⎝ m + 1, b (a + bv )⎟⎠ ⎥ ⎪
p
⎩ ⎣ ⎦⎭ ⎛5⎞ 3 ⎛3⎞ 1
function and Γ⎜ ⎟ = π & Γ⎜ ⎟ = π .
⎝2⎠ 4 ⎝2⎠ 2
[Re p > 0, b > 0]
∞
π (2 μu + 1) −uμ The fastest derivation uses a partial fraction
∫
x
e − μx dx = e [Re μ > 0] expansion and two integrals from Sect. 3.462,
x−u
3
u
2μ 2 below. A general approach will be outlined
with details left to the reader.
x a + bx a
• In = −
a + bx b b a + bx
(see Math. Summary, Partial Fractions)
set a = −u , b = 1 to get,
∞
⎡ u ⎤ − μx
I = ∫⎢ x−u + ⎥e dx
u ⎣ x − u ⎦
o In first integral apply integral,
Sect. 3.462, below,
∞

∫ (ax ± b ) e − px dx =
m

u
pb
±
a me a ⎛ pb ⎞
m +1
Γ⎜ m + 1, pu ± ⎟
p ⎝ a ⎠
with parameters
⎡ 1⎤
⎢⎣ p = μ , b = u , a = 1, m =
2 ⎥⎦
o In second integral apply integral,
Sect. 3.462, below,
∞ − px
e
∫u (ax ± b )n dx =
pb
±
n −1
p e ⎛a
pb ⎞
Γ⎜ − n + 1, pu ± ⎟
⎝ a ⎠
n
a
with parameters
⎡ 1⎤
⎢⎣ p = μ , b = u, a = 1, n = 2 ⎥⎦

—Sect. 3.363—
 111

¾ Verified on math.com
∞
x Exercise. Hint: Set m = −1, u = 0, v → ∞
∫ a + xe
− px
dx v

∫ x(a + bx ) e − px dx .
m
0 in above integral,
u
v
x Derive integral based on solution,
∫ (a + bx ) e − px dx v

∫ x (a + b )
m
e − px dx =
m
u

⎧⎡ ⎛ ⎞ ⎤ ⎫
⎪ ⎢ Γ ⎜ m + 2 , (a + bu )⎟ − ⎥
p
⎪
⎪⎢ ⎝ b ⎠ ⎥
− ⎪
⎪⎢ ⎛ ⎞ ⎥ ⎪
⎪ ⎢ Γ ⎜ m + 2 , (a + bv )⎟ ⎥
p
pa
⎪
e b
⎪⎣ ⎝ b ⎠ ⎦ ⎪
⎨ ⎬
p m + 2 b 1− m ⎪ ⎡ ⎛ ⎞ ⎤⎪
Γ m + 1, (a + bu )⎟ − ⎥
p
⎪ pa ⎢ ⎜⎝ b ⎠ ⎥⎪
⎪ ⎢ ⎪
⎪ b ⎢ ⎛ m 1, p (a bv )⎞ ⎥ ⎪
⎪ ⎢Γ ⎜ + + ⎟ ⎥⎪
⎩ ⎣ ⎝ b ⎠ ⎦⎭

by setting m → −1.
∞
ae − px 2 ap
[
e π 1 − Φ ap ( )] See below integral,
∫a+x dx =
[ ( )]
∞
2 x − u −qx π −qu
0 2 a
∫u x e dx = q e − uπ 1 − Φ qu
[Re p > 0, Re a > 0]
[Re q > 0]
3.363.1 Formula 3.363.1 in GR seems

[ ( )]
∞
x − u −qx π −qu straightforward, but it is a challenging
∫u x e dx = q e − uπ 1 − Φ qu integral. Its solution is well worth
consideration because of the logic and details
[Re q > 0] involved. The solution in Moll’s paper (on
error functions, p. 6) at
http://www.math.tulane.edu/~vhm/Table.html
Advanced solution by differentiation under the required four changes of variable,
integral sign, applying Form 3.462.12, Form. differentiation under the integral sign, and
⎛ 1 ⎞ other transformations. Patterned on Moll’s
2.33.19 and derived Γ⎜ ± , x ⎟ of Sect. 8.359, solution, a detailed derivation is given below.
⎝ 2 ⎠ Derivation—
below, all of which streamlines the work: 1. Change of variable:
x = u + t 2 ; dx = 2tdt
t = x − u ; t u = 0; t ∞ = ∞
Then,
∞
t2
I = 2e −qu ∫ 2 e −qt dt
2

0
t +u
o Use partial fraction expansion,

—Sect. 3.363—
 112

Treating u as constant, differentiating wrt q, t2 u

= 1− 2 ,
d
∞
x − u −qx
∞
x x − u −qx t +u
2
t +u
∫
dq u x
e dx = − ∫
u
x
e dx to give,
⎡ ∞ −qt 2 ∞
e − qt ⎤
2
∞
I = 2e ⎢ ∫ e dt − u ∫ 2
− qu
dt ⎥ =
= − ∫ x − u e −qx dx by Rule # 8 (Math. Summary) ⎢⎣ 0 t + u ⎥⎦
0
u
∞
π e −qt
2

⎡ ⎛ 3 ⎞ − 32 −qu ⎤ 1 e − 2ue ∫ 2
− qu − qu
dt
= − ⎢Γ⎜ ⎟q e ⎥ by Form. 3.462.13 with m = . q t + u
⎣ ⎝2⎠ ⎦ 2 0
where the first integral is solved by 3.326.2,
Int. back wrt q (apply Form. 2.33.19),
Γ(γ )
∞
m +1
⎛3⎞ −
3
⎛ 3 ⎞⎡
1
⎛ 1 ⎞⎤ ( )
∫0 x exp − βx dx = nβ γ , γ = n
m n

− Γ⎜ ⎟ ∫ q 2 e −qu dq = −Γ⎜ ⎟ ⎢− u 2 Γ⎜ − , qu ⎟⎥
⎝2⎠ ⎝ 2 ⎠⎣ ⎝ 2 ⎠⎦
⎛1⎞
Γ⎜ ⎟
πu ⎡ e − qu
⎛1 ⎞⎤ ⎛ 1 ⎞ Γ(γ ) ⎝2⎠ ,
= 2⎢ − Γ⎜ , qu ⎟⎥ by Γ⎜ − , x ⎟ to give, γ
= 2e − qu
2 ⎣⎢ qu ⎝2 ⎠⎦⎥ ⎝ 2 ⎠ nβ 1
2q 2
⎛3⎞ π ⎛1 ⎞ and the second integral requires more work.
and setting Γ⎜ ⎟ = & applying Γ⎜ , x ⎟,
⎝2⎠ 2 ⎝2 ⎠ Call that integral K.
⎡ e −qu
= πu ⎢ [ ⎤
− π 1 − Φ qu ⎥ ( )] 2. Change of variable on second integral,
⎢⎣ qu ⎥⎦ ∞
e − qt
2

K = 2ue ∫ 2
− qu
dt :
=
π
[
e −qu − uπ 1 − Φ qu ( )] 0
t +u
q r 2 = qt 2 ; r = t q ; dr = q dt
r2
r0 = 0; r∞ = ∞; t 2 =
q
Then,
∞
1 e −r
2

∫
− qu
K = 2ue dr
q r + qu
2
0
q
∞
e −r
2

= 2ue − qu
q∫ 2 dr
0
r + qu

3. Change of variable:
s = r 2 + qu; ω = qu
ds = 2rdr
s 0 = qu = ω ; s ∞ = ∞
r = s − qu = s − ω
Then,
e − ( s − qu ) ω = qu
∞

∫ 2rs ds =
− qu
K = 2u q e
qu = ω
∞ ∞
e− s e− s
u qe − qu qu
e ∫ ds = u q ∫ ds
ω s s −ω ω s s −ω

—Sect. 3.363—
 113

4. Change of variable:
s
s = ωy; ds = ωdy; y =
ω
yω = 1; y ∞ = ∞
ω
Then, noting that q= ,
u
∞
e − ωy
K = u q∫ dy =
1 ωy ωy − ω

ω∞ ωe −ωy ∞
e −ωy
u ∫1 ωy ω y − 1 ∫1 y y − 1dy
u dy = u

5. Call this last integral T . Differentiate T

under integral sign (wrt ω ) for the function,
e −ωy
f (ω , y ) = , to give,
y y −1
∂ ⎛⎜ e −ωy ⎞⎟
∞ ∞
e −ωy
u∫ dy = − u ∫ dy =
1
∂ω ⎜⎝ y y − 1 ⎟⎠ 1 y −1
π −ω
− u e (u, ω ≠ 0 )
ω
by either Formula 3.362.1,
∞
e − μx π −μ
∫1 x − 1 dx = μ e , or reduction formula
∞
e − px
above, ∫ dx .
u (a − bx )
n

6. Now integrate back (wrt ω ) to find T :

By Formula 2.33.19, Sect. 2.32, above, with

⎡ 1 1⎤
parameters ⎢m = n = β = 1, z = − ⎥ , then
⎣ 2 2⎦
−ω ∞ −r
e e dr
T =− u π∫ dω = u π ∫
ω ω r
⎛1 ⎞
= u π Γ⎜ , ω ⎟
⎝2 ⎠
by Form 3.462.19,
exp(− βx n ) Γ( z )
∞
1− m
∫u x m = nβ z , z = n ,
⎡ 1 ⎤
with parameters ⎢u = ω , m = z = , n = β = 1⎥
⎣ 2 ⎦

—Sect. 3.363—
 114

This, in turn, produces the error function,

⎛1 ⎞
u π Γ⎜ , ω ⎟ =
⎝2 ⎠
{ π [1 − Φ( ω )]} = [ ( )]
qu =ω
u π uπ 1 − Φ qu

π [1 − Φ ( x )], Sect.
⎛1 ⎞
by relation Γ⎜ , x ⎟ =
⎝2 ⎠
8.359, below. This completes the derivation
for the second integral term, K.

7. Combine above step 6 result with step 1

result to get published solution,

[ ( )]
∞
x − u −qx π −qu
∫u x e dx = q e − uπ 1 − Φ qu
¾ Verified on math.com
NOTE: Solve 3.363.2 using similar
approach, noting the partial fraction
expansion,
1 1 a
= − .
x x−a x−a x x−a
∞
x − u − qx
NOTE: Can ∫
u
x
e dx be generalized by

v
(a + bx )
m

∫
u
x
e − qx dx ? Math.com unable to solve

3.363.1 with limits u, v and the indefinite

integral.

—Sect. 3.363—
 115

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
∞ 1 Using Formula 3.326.2 with parameters,
n−
In Formula 3.371, ∫x
0
2
e − μx dx, 1 3
[m = n + , n = 1, β = μ , γ = n + ]
2 2
Add Formula,
⎛ 1⎞
Γ⎜ n + ⎟
= ⎝
2⎠
1
n+
μ 2

⎛ 3⎞ Using Formula 3.326.2 with parameters,

∞ 1 Γ⎜ n + ⎟ 1 3
dx = ⎝ 3 ⎠ =
n+ 2 [m = n + , n = 1, β = μ , γ = n + ] ,
∫x 2 − μx
e
n+ 2 2
0
μ 2
⎛ 3 ⎞
Γ⎜ n + ⎟
(2n + 1)(2μ )−n−1 (2n − 1)!! π
∞ 1
I = x 2 e − μx dx = ⎝ ⎠ . The numerator can be
n+ 2
μ ∫ n+
3
0
μ 2
[Re μ > 0, n ≥ 0] expressed as,
⎛1 ⎞ ⎛1 ⎞ ⎛ 2n + 1 ⎞⎛⎜ π (2n − 1)!! ⎞⎟
⎜ + n ⎟Γ⎜ + n ⎟ = ⎜ ⎟
⎝2 ⎠ ⎝2 ⎠ ⎝ 2 ⎠⎜⎝ 2n ⎟
⎠
⎛1 ⎞
by Formula 8.339 2, Γ⎜ + n ⎟ .
⎝2 ⎠
Thus,

I=
(2n + 1) π (2n − 1)!!
3
n+
2 n+1 μ 2
π
= (2n + 1)(2 μ ) (2n − 1)!!
− n −1
, n > 0.
μ
¾ Verified on Math.com

——Sect. 3.371——
 116

⎛3 ⎞ ⎛3 ⎞
Γ⎜ − n ⎟
∞ Γ⎜ − n ⎟ ∞ 1
−n
⎝ 2 ⎠
∫x
− μx
1 dx =
dx = ⎝ 3 ⎠
2
−n 2 e
∫x 2 − μx 3
−n
e 0
μ2
−n
0
μ2 ⎡ 1 3 ⎤
⎢m = 2 − n, β = μ , n = 1, γ = 2 − n ⎥.
⎣ ⎦
=
(− 1)n+1 (2n − 1)(2μ )n−1 π by Formula 3.362.2 with parameters,
(2n − 1)!! μ
⎛3 ⎞ ⎡⎛ 1 ⎞ ⎤ ⎛1 ⎞ ⎛1 ⎞
[Re μ > 0, n ≥ 0] Γ⎜ − n ⎟ = Γ ⎢⎜ − n ⎟ + 1⎥ = ⎜ − n ⎟Γ⎜ − n ⎟,
⎠ ⎝2 ⎠
⎝2 ⎠ ⎣⎝ 2 ⎠ ⎦ ⎝2
where,
⎛1 ⎞ (− 1) π 2
n n
Γ⎜ − n ⎟ = by Formula 8.339 .3.
⎝2 ⎠ (2n − 1)!!
Thus,

I=
(− 1) (1 − 2n ) π 2 n−1 n− 2
n
μ
3

(2n − 1)!!
=
(− 1)n+1 (2n − 1) π 2 n−1 μ n− 32
(2n − 1)!!
=
(− 1)n +1 (2n − 1)(2μ )n −1 π
(2n − 1)!! μ
NOTE: unreduced expression,
(2n −1) = 1
(2n −1)!! (2n - 3)!!
allows n = 0 calculation.

¾ Verified on math.com

——Sect. 3.371——
 117

∞ n−
p
• Apply Formula 3.362,2 with parameters
∫x e − μx dx =
q
⎡ p p ⎤
0 ⎢m = n − q , n = 1, β = μ , γ = n − q + 1⎥ , to give:
⎛ ⎣ ⎦
p ⎞
Γ⎜⎜ n − + 1⎟⎟ ⎛ p ⎞ ⎛ p⎞
⎝ q ⎠= Γ⎜⎜ n − + 1⎟⎟ Γ⎜⎜ n − ⎟⎟
q ⎠ nq − p ⎝
I= ⎝
q⎠
=
p
n − +1
μ q p
n − +1 q p
n − +1
μq q
μ
n ⎛ p⎞
∏ (kq − p ) Γ⎜⎜ m − ⎟⎟ p −n−1 • Apply from Sect. 8.339, below,
⎝ q⎠ q
μ ∏ (qk − p )
k =1 n
m −1 n − m +1
⎛ p⎞ ⎛ p⎞
∏ (kq − p )
q Γ⎜ n − ⎟ = k =1
Γ⎜⎜ m − ⎟
k =1
⎝ q ⎠ q n − m (nq − p ) m −1 (qk − p ) ⎝
∏k =1 q ⎟⎠
⎡ p ⎤ ⎡p ⎛ p⎞ ⎤
⎢Re μ > 0, q not a natural number,⎥ ⎢ q not a natural number, m = int ⎜⎝ q ⎟⎠ + 1, p > q ⎥
⎢ ⎥ ⎣ ⎦
⎢ ⎛ p⎞ ⎥ to give:
⎢m = int ⎜⎜ ⎟⎟ + 1, n ≥ 0 ⎥ n ⎛ p⎞
⎝q⎠
⎣ ⎦ ∞ n−
p ∏ (kq − p ) Γ⎜⎜⎝ m − q ⎟⎟⎠ p
− n −1
I =∫x q μx
e dx = k =1
m −1
μ q

q n−m+1
0
∏ (kq − p )
k =1

p 1
• Test: Let = . Then m=1, and
q 2
n

∞
n−
1 ∏ (2k − 1) ⎛ 1 ⎞ − −n
1
I =∫x 2
e − μx
dx = k =1
Γ⎜ ⎟ μ 2
0
2n ⎝2⎠

=
(2n − 1)!! πμ
1
− −n
2
2n
which is Formula 3.371.1.

——Sect. 3.371——
 118

∞ n−
p

• ∫x e μx dx ,
q
⎛ p ⎞ Similar to above integral,
∞ n+ p
Γ⎜⎜ n + + 1⎟⎟ 0
⎝ q ⎠=
∫ x e dx =
q − μx employing Sect. 8.339 formula, below,
p
n + +1
∏ ( p − qk ) ∏ ( p + kq − q ) ⎛ p
m n
μ
0 q
⎛ p⎞ ⎞
n Γ⎜ n + ⎟ = k =1 k =1
Γ⎜⎜ − m ⎟⎟
nq + p m
( ) (
∏ p-kq ∏ p + kq − q ) ⎝ q⎠ qm + n ⎝q ⎠
n + m +1 k =1 k =1
q
⎡p ⎛ p⎞ ⎤
⎛p
p
⎞ −n − q −1 ⎢ q not a natural number, m = int ⎜⎝ q ⎟⎠, p > q ⎥
Γ⎜ − m ⎟ μ
⎣ ⎦
⎝q ⎠ p 1
• Test: Let = . Then m = 0, and
⎡ p ⎤ q 2
⎢Re μ > 0, q not a natural number,⎥
⎢ ⎥
∞
I =∫x
n+
1
2
e − μx dx =
(2n + 1)(2n − 1)!! πμ
−n−
3
2
⎢ ⎛ p⎞ ⎥ 2 n+1
⎢m = int ⎜⎜ ⎟⎟, n ≥ 0
0
⎥
⎣ ⎝q⎠ ⎦ π
= (2n + 1)(2 μ ) (2n − 1)!!
− n −1
,
μ
consistent with above integral.
∞ p
−n ∞ n−
p

∫x ∫x e μx dx ,
− μx
q
e dx = • Similar to above integral, q

0 0

⎛p ⎞ employing Sect. 8.339 formula, below,

Γ⎜⎜ − n + 1⎟⎟
(− 1)m + n q n − m ∏ (qk − p ) ⎛ p ⎞
m
⎝q ⎠= ⎛ p ⎞
Γ⎜ − n ⎟ = k =1
Γ⎜⎜ − m ⎟⎟
( p)
p
− n +1
⎝ ⎠ ∏ −
n

μq
q qk ⎝q
k =1 ⎠
m ⎡p ⎛ p⎞ ⎤
∏ (kq − p ) ⎢ q not a natural number, m = int ⎜⎝ q ⎟⎠, p > q ⎥
⎣ ⎦
(− 1)m+ n ( p − nq )(q n−m−1 ) kn=1 p 1
∏ (kq − p )
k =1
• Test: Let = . Then m = 0 , and
q 2
⎛p ⎞
p
n − −1
∞
I =∫x
1
−n
− μx
dx =
(− 1) (1 − 2n )2 n −1 ⎛ 1 ⎞ n− 2
n
Γ μ
3

Γ⎜⎜ − m ⎟⎟ μ q
2
e ⎜ ⎟
⎝2⎠
n
⎝q ⎠ 0
∏ (2k − 1)
k =1
⎡ p ⎤
⎢Re μ > 0, q not a natural number,⎥ π
= (− 1) (2n − 1)(2μ )n−1 (2n − 1)!!
n +1
⎢ ⎥ ,
⎢ ⎥
μ
⎛ p⎞
⎢ m = int ⎜⎜ ⎟⎟, n ≥ 0 ⎥ consistent with above integral.
⎣ ⎝q⎠ ⎦

——Sect. 3.371——
119
 120

FORMULA PROOF OUTLINE

(formula references are to G & R
7e)
Formula 3.381.8 ds
In integral, set s = β x n , = nβ x n −1 ,
m +1
( ) γ (ν , βu )
u
−1 dx
∫ x e dx = nβ
m − βx ν
ν=
n
n
βu n
n m +1 1
∫s
ν −1 − s
ν= to get, I =
0
e ds
[ u > 0, Re ν > 0, Re n > 0, Re β > 0 ] n nβ ν 0

and apply definition of

lower incomplete gamma function,
γ (α , x ), Fomula 8.350.1
Formula 3.381.9 ds
∞ In integral, set s = βx n , = nβx n −1 ,
m +1
∫x
m − βx n
e ( ) Γ(ν , βu )
dx = nβ ν
−1 n
ν=
dx
∞
n m +1 1
nβ ν β∫u n
u
ν = to get, I = sν −1e − s ds
[Re n > 0, Re β > 0 ] n
and apply definition of
upper incomplete gamma function,
Γ(α , x ) of Formula 8.350.2.
Formula 3.381.10 Apply functional relation of complete
∞ u ∞
gamma function Γ(α ) to lower
∫ x e dx = ∫ x e dx + ∫ x e dx =
m − βx n
m − βx m − βx n n

0 0 u
incomplete gamma function γ (α , x )

(nβ ) γ (ν , βu ) + (nβ ) Γ(ν , βu )

v −1 n v −1 n and upper incomplete gamma function
Γ(α , x ), Formula 8.356.3
m +1
ν= [u > 0, Re ν > 0, Re n > 0, Re β > 0 ]
n
See also Formula 3.326 2
u Derived from Formula 3.462.15,
∫ x e dx .
ρ −1 − x
Formula 3.381.2,
( ) γ (ν , βu )
u
−1
∫x
m − βx n
0 e dx = nβ ν n

Add Formula 0

= γ (ρ , u ) ν=
m +1
[ u > 0, Re ν > 0, Re n > 0,
n
Re β > 0 ]
with parameters,
[m = ρ − 1, β = n = 1,ν = ρ ]
∞
e−x Change existing integral to include
Formula 3.381.6, ∫u xν dx result,
Γ(1 −ν , u ) , derived from previously
Add Formula
submitted formula,
= Γ(1 −ν , u )

——Sect. 3.381——
 121

∞
e − βx
n
Γ z, βu n ( )
∫u x m dx =
nβ z
,

1− m
z= [Re β > 0, Re n > 0]
n
with parameters,
[m = ν , β = n = 1, z = 1 −ν ]
Γ(ν , μ u ) − Γ(ν , μ v ) Apply
v

∫x
ν −1 − μx
e dx =
( )
v

u
μν ∫x
m
exp − β x n dx =
[Re μ > 0] u

( ) (
Γ γ , βu n − Γ γ , βv n
,γ =
m +1 )
γ
nβ n
[Re β > 0, n > 0]
with parameters,
[m = ν − 1, n = 1, β = μ , γ = ν ]
NOTE: Formulas 3.381 1 through
3.381.4 follow by substituting u, v.
Γ(γ )
∫
∞
(
x 2 m exp − β x 2 n dx =) 2m, 2n are assumed even-number
−∞ nβ γ integers or fractions with even-integer
numerators;
2m + 1
γ= [re β > 0, re n > 0, re γ > 0] thus, integrand is even function,
2n ds 2 n −1
Set s = β x = 2 nβ x
2n
, ,
dx
1 2n
x = ⎛⎜ s ⎞⎟
2 m +1
,γ = 2n > 0
⎝β ⎠
and substitute . Integral results.

⎛ x−a⎞ ⎡ ⎛ x − a ⎞ 2n ⎤ bΓ(γ ) 2m, 2n are assumed even-number

2m
∞
∫−∞ ⎜⎝ b ⎟⎠ exp ⎢− β ⎜ ⎟ ⎥ dx =
⎢⎣ ⎝ b ⎠ ⎥⎦ nβ γ
integers or fractions with
even-integer numerators; thus,
2m + 1 integrand is even function.
γ = ⎛ x − a ⎞ ds 1
2n Set s = ⎜ ⎟, = and substitute.
⎝ b ⎠ dx b
[re β > 0, re n > 0, re b > 0,
dt 2 n −1
re γ > 0, - ∞ < re a < ∞]. Then, set t = βs = 2 nβ s
2n
, ,
ds
1 2n
⎛t⎞ 2m + 1
s=⎜ ⎟ ,γ = > 0 and substitute.
⎝β ⎠ 2n
Integral results.

——Sect. 3.381——
 122

⎛ a ⎞ In integral,
∞ Γ⎜ γ + ⎟
n+b⎠
dx = ⎝
set
∫x
n+b
m +α − βx
e a
m +1
= (n + b )βx n+b−1 , γ =
γ+ ds
0
(n + b )β n +b s = β x n +b ,
dx n+b

m +1 ⎡ a ⎤ to get,
γ= ⎢ Re β > 0, Re n + b > 0, Re γ + > 0⎥
n+b ⎣ n+b ⎦ 1
∞ ⎛
⎜γ +
α ⎞
⎟ −1

∫s ⎝ n +b ⎠ −s
I= α
e ds
γ+
(n + b )β n +b 0

⎛ a ⎞ Similar to
∞ Γ⎜ γ + ⎟
⎝ n−b⎠ ⎛ a ⎞
∫
m +α − βx n −b
x e dx = ∞ Γ⎜ γ + ⎟
n+b⎠
dx = ⎝
a

∫x
n+b
γ+ m +α − βx
0
(n − b )β n −b e a
γ+
m +1 ⎡ a ⎤
0
(n + b )β n +b

γ= ⎢ Re β > 0, Re n − b > 0, Re γ + > 0⎥

n−b ⎣ n−b ⎦
⎛ a ⎞ Similar to
∞ Γ⎜ γ − ⎟
n−b⎠ ⎛ a ⎞
∫ dx = ⎝
n −b
x m−α e − βx ∞ Γ⎜ γ + ⎟
a
⎝ n+b⎠
∫
γ− m +α − βx n + b
0
(n − b )β n −b x e dx = a
γ+
m +1 a
0
(n + b )β n +b

γ= [Re β > 0, Re n − b > 0, Re γ − > 0]

n−b n−b
⎛ a ⎞ Similar to
γ ⎜ν + , β u n +b ⎟
⎛ a ⎞
u
⎝ n+b ⎠
∫
n+b
x m+α e − βx dx = ∞ Γ⎜ γ + ⎟
a
⎝ n + b⎠
∫x
n+b
ν+ m +α − βx
0
(n + b )β n +b e dx = a
γ+
m +1
0
(n + b )β n +b

ν=
n+b
a
[u > 0, Re β > 0, Re n + b > 0, Reν + > 0]
n+b
⎛ a ⎞ Similar to
γ ⎜ν − , β u n+b ⎟
⎛ a ⎞
u
n+b
∫x dx = ⎝ ⎠
n+b
m−α − βx
e ∞ Γ⎜ γ + ⎟
a
⎝ n+b⎠
∫
ν− m +α − βx n + b
0
(n + b )β n +b x e dx = a
γ+
m +1
0
(n + b )β n +b

ν=
n+b
a
[u > 0, Re β > 0, Re n + b > 0, Reν − > 0]
n+b

——Sect. 3.381——
 123

⎛ a ⎞ Similar to
γ ⎜ν + , β u n−b ⎟
⎛ a ⎞
u
n−b
∫ x m+α e −βx dx = ⎝ ⎠
n −b
∞ Γ⎜ γ + ⎟
a
⎝ n+b⎠
∫
ν+ m +α − βx n + b
0
(n − b )β n −b x e dx = a
γ+
m +1
0
(n + b )β n +b

ν=
n−b
a
[u > 0, Re β > 0, Re n − b > 0, Reν + > 0]
n−b
⎛ a ⎞ Similar to
γ ⎜ν − , β u n−b ⎟
⎛ a ⎞
u
n−b
∫ x m−α e − βx dx = ⎝ ⎠
n−b
∞ Γ⎜ γ + ⎟
n+b⎠
dx = ⎝
a

∫x
n+b
ν− m +α − βx
0
(n − b )β n −b e a
γ+
m +1
0
(n + b )β n +b

ν=
n−b
a
[u > 0, Re β > 0, Re n − b > 0, Reν − > 0]
n+b
⎛ a ⎞ Similar to
∞ Γ⎜ν + , βu n+b ⎟
⎝ n+b ⎠ ⎛ a ⎞
∫
n+b
x m+α e − βx dx = ∞ Γ⎜ γ + ⎟
+
dx = ⎝
a n b⎠
∫x
n+b
ν+ m +α − βx
u
(n + b )β n +b e a
γ+
m +1
0
(n + b )β n +b

ν= [Re β > 0, Re n + b > 0]

n+b
⎛ a ⎞ Similar to
∞ Γ⎜ν − , β u n +b ⎟
n+b ⎛ a ⎞
∫ x m−α e − βx dx = ⎝ ⎠
n+b
∞ Γ⎜ γ + ⎟
a
⎝ n+b⎠
∫
ν− m +α − βx n + b
u
(n + b )β n +b x e dx = a
γ+
m +1
0
(n + b )β n +b

ν= [Re β > 0, Re n + b > 0]

n+b
⎛ a ⎞ Similar to
∞ Γ⎜ν + , β u n−b ⎟
n−b ⎛ a ⎞
∫ x m+α e −βx dx = ⎝ ⎠
n −b
∞ Γ⎜ γ + ⎟
+
dx = ⎝
a n b⎠
∫x
n+b
ν+ m +α − βx
u
(n − b )β n−b e a
γ+
m +1
0
(n + b )β n +b

ν= [Re β > 0, Re n − b > 0]

n−b
⎛ a ⎞ Similar to
∞ Γ⎜ν − , βu n−b ⎟
n−b ⎛ a ⎞
∫ x m−α e − βx dx = ⎝ ⎠
n−b
∞ Γ⎜ γ + ⎟
a
⎝ n+b⎠
∫
ν− m +α − βx n + b
u
(n − b )β n −b x e dx = a
γ+
m +1
0
(n + b )β n +b

ν= [Re β > 0, Re n − b > 0]

n−b

——Sect. 3.381——
 124

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
Formula 3.382.3, Derived from Formula 3.462.15,
∞ pb
±

∫ (1 + x )
−ν − μx ∞ − px n −1
e dx . e p e a
⎛ pb ⎞
0 ∫ (ax ± b )
0
n
dx =
an
Γ⎜ − n + 1,±
⎝ a ⎠
⎟
Add Formula
⎡ ⎤
= μ ν −1e μ Γ(−ν + 1, μ )
b
⎢ p > 0, arg a < π ⎥
⎣ ⎦
with parameters, [ρ = μ , a = b = 1, n = ν ]

——Sect. 3.382——
125
 126

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
0
⎡1 1 ⎤ x Derived in Sect. 8.367, below.
∫ ⎢⎣ x − e
−∞
x
− 1⎥⎦
e dx = C

—Sect. 3.427—
 127

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
p p ⎛ ρ⎞
∞ − μx t Γ⎜ − ⎟
− vx t
−e μ −v ⎛t − ρ ⎞ Solve ∞
e − vx by Formula 3.362.2 with parameters
t
t t
e ⎝ t ⎠
∫0 x ρ +1 ρ
Γ⎜dx =
⎝ t ⎠
⎟ ∫0 x ρ +1 dx = − ρ
tv t

⎡ ρ ⎤ ⎡ ρ⎤
⎢ Re μ > 0, Re v > 0, Re ρ > 0, Re < 1⎥ ⎢⎣m = − ρ − 1, β = v, n = t , γ = − t ⎥⎦ , and solve
⎣ t ⎦
⎛ ρ⎞ t ⎛t −ρ ⎞
Γ⎜ − ⎟ = − Γ ⎜ ⎟ by Formula 8.331,
⎝ t ⎠ ρ ⎝ t ⎠
ρ
Γ(α + 1) = αΓ(α ), with α = − .
t
⎛ ρ⎞
Γ⎜ − ⎟
Solve ∞
e − μx
t

⎝ t ⎠ in a similar manner. Thus,

∫0 x ρ +1 dx = − ρ
tμ t

⎛t − ρ ⎞ ⎛t − ρ ⎞ ρ ρ
ρ Γ⎜ ⎟ ρ Γ⎜ ⎟
t ⎝ t ⎠ t ⎝ t ⎠ μ −v t ⎛t − ρ ⎞
t
I = −v t +μt = Γ⎜ ⎟
ρ t ρ t ρ ⎝ t ⎠
NOTE: if t = 1, Formula 3.434.1 follows
¾ Verified on math.com
Formula 3.434.2 Our proof separates the terms, uses Leibnitz’s Rule #6 in Math.
exp(− ax ) − exp(− bx )dx
∞ Summary to find derivatives, then integrates back.
b
∫ x
= ln
a
Whittaker/Watson suggest another way by two double integrals.
0 Thus, let separate integrals be called I = I (a ) − I (b)
[Re a > 0, Re b > 0] ∞
exp(− ax )dx
I (a) = ∫ .
0
x
∂ ⎛ exp(− ax ) ⎞ exp(− ax )dx
∞ ∞
I ′(a ) =
dI (a )
=∫ ⎜ ⎟dx = ∫ − x
da 0
∂a ⎝ x ⎠ 0
x
1
=− .
a
dI (a ) 1
Then,
da∫ = − ∫ da = − ln a = I (a ).
a
∂ ⎛ exp(− bx ) ⎞
∞
1
Likewise, I ′(b) = ∫ ⎜ ⎟dx = − .
0
∂b ⎝ x ⎠ b
dI (b) 1
∫ db
= − ∫ db = − ln b = I (b)
b
The constant of integration C is 0 by, I = I (a ) − I (b),
I (0) = 0; I (0) = ln(0) − ln(0) + C ⇒ C = 0.
exp(− ax ) − exp(− bx )dx b
Thus, = − ln a − (−) ln b = ln
x a
NOTE: Integral is called a type of “Frullani Integral”.

—Sect. 3.434—
128
 129

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
∞ Γ ()
n
⎛n⎞
Γ⎜ ⎟
∫x
n −1 − ρx 2
e dx = 2
[ ρ > 0, n ≥ 1] ∞
I = ∫ x n −1e − ρx dx = ⎝ n ⎠ by Formula 3.326.2
n 2 2
0
2ρ 2
0
2ρ 2
2π (n − 2)!!
= [ρ > 0, n = 1,3,K] with parameters,
2(2 ρ )
n
n
2
[m = n − 1, n = 2, β = ρ , γ = ] . The first
2
form can be used generally for any n ≥ 1.
⎛n⎞
∞ Γ⎜ ⎟
If n is odd, I = ∫ x n−1e − ρx2 dx = ⎝ 2n⎠ can be
0
2ρ 2
further simplified, by noting that,
⎛n⎞ 2
1− n
π (n − 1)! . This standard form
Γ⎜ ⎟ =
⎝2⎠ ⎛ n −1⎞
⎜ ⎟!
⎝ 2 ⎠
can be expressed more compactly as
⎛n⎞ π (n − 2 )!!
Γ⎜ ⎟ = n −1
, as shown below, Sect.
⎝2⎠ 2
2
8.339.

Then, I =
( )⎛
π n − 2 !! ⎜ 1
⎞
⎟= ( )
2π n − 2 !!
.
n −1
⎜ 2⎟
2 (2 ρ ) 2
n n
2 2 ⎝ 2ρ ⎠
¾ Verified on Math.com
∞
(n − 1)! Follows from Formula 3.362.2 with
∫x
2 n −1 − ρx 2
e dx = [ ρ > 0, n = 1,2, K] parameters,
2ρ n
0
[m = 2n − 1, n = 2, β = ρ , γ = n] ,
⎛ 2n − 1 + 1 ⎞
Γ⎜ ⎟
I= ⎝ 2 ⎠ = Γ(n )
2 n −1+1
2ρ n
2ρ 2
¾ Verified on Math.com

——Sect. 3.461——
 130

⎛ 1⎞ Follows from Formula 3.362.2 with

∞ Γ⎜ n − ⎟ parameters,
2 ( n −1) − ρx 2 ⎝ 2⎠
∫ x e dx = ⎡ 1⎤
⎢m = 2n − 2, n = 2, β = ρ , γ = n − 2 ⎥ ,
1
n−
0
2ρ 2
⎣ ⎦
ρπ (2n − 3)!! ⎛ 1⎞
= [ρ > 0, n = 1 ∞ Γ⎜ n − ⎟
(2 ρ )n I = ∫ x 2(n−1)e − ρx
2
dx = ⎝
n−
2⎠
1
.
0
2ρ 2
The author has derived, as shown below,
Sect. 8.339,
n −1
π ∏ (2k − 1)
⎛ 1⎞
Γ⎜ n − ⎟ = k =1
=
⎝ 2⎠ 2 n −1
π (1)(3)(5)K (2n − 3)
2 n −1
π (2n − 3)!!
= .
2n −1
Then,
∞
π (2n − 3)!!
I = ∫ x 2(n −1)e − ρx = =
2

⎛ n − 12 ⎞
0
2 n −1 ⎜ 2ρ ⎟
⎜ ⎟
⎝ ⎠
πρ (2n − 3)!!
(2 ρ )n
¾ Verified on Math.com

− ρ 2 x2 In Formula 3.461.1,
∞ e
dx =
(− 1) π 2n −1 ρ 2 n −1
n
∞ − ρ 2 x2
∫ e
0 x 2n (2n − 1)!! ∫u x 2n dx =
[ρ > 0, n = 0,1,2,...]
⎛1 ⎞
Γ⎜ − n, ρ 2 u 2 ⎟
⎝2 ⎠,
2ρ 1− 2 n

let u = 0.
¾ Or apply previously derived,
Γ( z )
∞ − βx n
e
∫0 x m dx = nβ z ,
1− m 1
z= , z = − n,
n 2
⎛1 ⎞
and use Γ⎜ − n ⎟, Formula 8.339.3.
⎝2 ⎠
¾ verified on math.com

——Sect. 3.461——
 131

∞ π Apply Formula 3.362.2 to even function,

[ρ > 0]
2
− ρx
∫e dx = ∞
ρx 2 dx = π
−∞ ρ I = 2∫ e
0
−

¾ verified on math.com
∞
2 n − ρx
dx =
2 (2n − 1)!! π [ρ > 0]
Since integrand is an even function, multiply
∫x e
−∞ (2 ρ )n ρ Formula 3.461.2 by 2.

v
∫e
− qx
2 dx 1 − qu2 1 − qv2
= e − e [ ( ) ( )]
+ qπ Φ v q − Φ u q
⎛ 1 ⎞
Γ⎜ − , qx2 ⎟ v
=− ⎝ ⎠ =
u x2 u v v
− qx
2 dx 2
∫e
[Re q > 0] u x2 −
1

2q 2 u
⎛ 1 ⎞ ⎛ 1 ⎞
Γ⎜ − , qu2 ⎟ − Γ⎜ − , qv2 ⎟
⎝ 2 ⎠ ⎝ 2 ⎠
1
−
2
2q
from indefinite integral solution, previously
submitted,
e − βx
n
β z Γ − z, βx n ( )
∫ x m dx = − n
z ∞
β e −t
m −1
=− ∫
n βx n t z +1
z=
n
⎡ 1⎤
with parameters ⎢ β = q, m = n = 2, z = ⎥
⎣ 2⎦

⎛ 1 ⎞ ⎛ 1 ⎞
Γ⎜ − , qu 2 ⎟ − Γ⎜ − , qv 2 ⎟
Then ⎝ ⎠ ⎝ 2 ⎠ is solved
2
1
−
2q 2
by formula in Sect. 8.359, below,
⎛ 1 ⎞ ⎡ e− x ⎛ 1 ⎞⎤
Γ⎜ − , x ⎟ = 2 ⎢ − Γ⎜ , x ⎟ ⎥ ,
⎝ 2 ⎠ ⎣ x ⎝ 2 ⎠⎦
⎛ 1 ⎞
Γ⎜ − , qu 2 ⎟
so that ⎝ ⎠=
2 1 − qu2 ⎛1 ⎞
1 e − qΓ ⎜ , qu2 ⎟
− u ⎝2 ⎠
2q 2
and,
⎛ 1 ⎞
Γ⎜ − , qv 2 ⎟
⎝ 2 ⎠ = 1 − qv2 ⎛1 ⎞
1 e − qΓ⎜ , qv2 ⎟.
− v ⎝2 ⎠
2
2q
Applying formula in Sect. 8.359,
⎛1 ⎞
Γ⎜ , x ⎟ = π − π Φ x , ( )
⎝2 ⎠
leads to integral.
NOTE: If v → ∞, Φ (∞ ) = 1, then
——Sect. 3.461——
 132

∞
∫e
u
− qx
2 dx 1 −qu2
[ ( )]
= e + qπ 1 − Φ u q
x2 u
of Formula

4.461.5

——Sect. 3.461——
 133

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
v ¾ Extension of previously submitted
∫ x e dx =
n+b
m ± a − βx
formulas,
u ∞ u

∫ x e dx, ∫ x e dx,
n ±b n ±b
m ± a − βx m ± a − βx
⎛ a ⎞ ⎛ a ⎞
Γ⎜ γ ± , β u n + b ⎟ − Γ⎜ γ ± , βv n +b ⎟
⎝ n+b ⎠ ⎝ n+b ⎠
0 0
.
∞

∫x
a n ±b

(n + b )β
γ±
n +b
m± a
e − βx dx.
u
m +1
γ = [Re β > 0, Re n + b > 0] Derivation—
n+b • Solve integral separately for
numerator terms
x m + a & x m − a and assign the signs
for general expression.
• In integral, set
s = β x n + b , = (n + b )β x n + b −1 ,
ds
dx
1
⎛ s ⎞ n +b m +1
x = ⎜⎜ ⎟⎟ , γ = ,
⎝β ⎠ n+b
su = β u n + b , sv = βv n + b ,
to obtain a gamma-function structure,
sv ⎛ a ⎞
1 ⎜γ ± ⎟ −1

∫s
n+b ⎠ −s
I= a
⎝
e ds .
γ±
(n + b )β n + b su

• Use same indefinite-integral

approach outlined in Sect. 3.323
for derivation of
v
− (ax )dx ⇒ (
Γ γ±
a
)(
, s −Γ γ ±
a
,s ),
∫e
2
+ kbx + c n +b u n +b v
a
γ±
u ( n +b ) β n +b

and apply definition of upper

incomplete gamma
function Γ(α , x ), Formula 8.356.3.
¾ Integral verified on
Math.com

Factor e ± a and apply Formula

Formula 3.462.9 3.326.2
∞
e ±a ⎛1⎞
∫ e − βx ±a
[Re β > 0, Re n > 0]
n
dx = Γ⎜ ⎟
⎝n⎠
1
0
nβ n

—Sect. 3.462—
 134

e ±a ⎛1 ⎞
u

∫e
− βx ±a
dx = γ ⎜ , βu n ⎟
n

Similar to
⎝n ⎠
1
0
nβ n ∞
e ±a ⎛1⎞
∫ e − βx
n
±a
dx = Γ⎜ ⎟
[Re n > 0, Re β > 0] 1
⎝n⎠
0
nβ n

∞ ±a
e ⎛1 ⎞
∫e
− βx n ± a
dx = Γ⎜ , β u n ⎟ Similar to
⎝n ⎠
1
u
nβ n ∞
e ±a ⎛1⎞
∫ e − βx
n
±a
dx = Γ⎜ ⎟
[Re n > 0, Re β > 0] 1
⎝n⎠
0
nβ n

∞ a
⎛2⎞ ⎛1⎞
e In integral, set s = β x n − a,
ds
= nβx n −1 ,
∫ ( x + b )e
−βxn +a a
dx =
Γ⎜ ⎟ + be 1 Γ⎜ ⎟
⎝n⎠ ⎝n⎠
2
dx
0
nβ nβ n n
1
[Re n > 0, Re β > 0] ⎛s+a⎞ n
x = ⎜⎜ ⎟⎟ ,
⎝ β ⎠
separate integrand expressions and
apply definition of complete gamma
function, Γ(α ), formula 8.310.
∞ a
⎛ 2 ⎞ bea ⎛ 1 ⎞
e
∫ ( x − b)e − βx
n
+a
Γ dx =
⎜ ⎟− Γ⎜ ⎟ Similar to above
⎝ ⎠ ⎝n⎠
2 1
n
0
nβ n nβ n
[Re n > 0, Re β > 0]
v Proof and approach similar to above
∫ x e dx =
n −b
m ± a − βx
integral,
u

⎛ a ⎞ ⎛ a ⎞ ⎛ a ⎞ ⎛ a ⎞
Γ⎜ γ ± , β u n − b ⎟ − Γ⎜ γ ± , β v n −b ⎟ v Γ⎜ γ ±
⎝ n+b
, β u n + b ⎟ − Γ⎜ γ ±
⎠ ⎝ n+b
, βv n + b ⎟
⎠
n−b n−b ∫x
m ± a − βx n + b
dx =
⎝ ⎠ ⎝ ⎠ e
γ±
a

a
u
(n + b)β n+b

γ±
(n − b )β n −b γ=
m +1
n+b
[Re β > 0, Re n + b > 0]
m +1
γ = [Re β > 0, Re n − b > 0]
n−b
v
e − βx
n+b
¾ See below
∫u x m± a dx = ¾ Extension of previously submitted
formulas,
⎛ a ⎞ ⎛ a ⎞ ∞ − βx n u − βx n ∞ − βx n
Γ⎜ z m , β u n + b ⎟ − Γ⎜ z m , βv n +b ⎟ e e e
⎝ n+b ⎠ ⎝ n+b ⎠ ∫0 x m dx, ∫0 x m dx, ∫u x m dx .
a
(n + b )β
zm
n+b

1− m
z= [Re β > 0, Re n + b > 0]
n+b

—Sect. 3.462—
 135

• Solve integral separately for

denominator terms
x m + a & x m − a and assign the signs
for general expression.
• In integral, set

= (n + b )βx n + b −1 ,
ds
s = βx n +b ,
dx
1

⎛s⎞ 1− m
n+b
x =⎜ ⎟ ,z = ,
⎝β ⎠ n+b
su = β u n + b , sv = β v n + b ,
to obtain a gamma-function structure,
sv ⎛ a ⎞
1 ⎜ zm ⎟ −1

∫s ⎝ n +b ⎠ −s
I= a
e ds .
(n + b )β
zm
n +b su

• Use same indefinite-integral

approach outlined in Sect. 3.323
for derivation of
(
v
)dx ⇒ (
Γ zm
a
)(
, s −Γ z m
a
,s ),
∫e
− ax 2 + kbx + c n +b u n +b v
a
zm
u ( n +b )β n +b

and apply definition of upper

incomplete gamma
function Γ(α , x ), Formula 8.356.3.
NOTE: following seven (7) similar
integrals result by substituting values
for u, v, su & sv by noting that, for the
incomplete gamma function, Γ(α , x ) :
Γ(a,0 ) = Γ(a ), Γ(a, ∞ ) = 0, and
Γ(a ) = Γ(a, x ) + γ (a, x ).

¾ Integral verified on
Math.com

n −b
e − βx Proof and approach similar to above
v

∫u x m±a dx = integral,
v − βx n + b
e
⎛ a ⎞ ⎛ a ⎞ ∫ dx.
xm ± a
Γ⎜ z m , β u n − b ⎟ − Γ⎜ z m , β v n −b ⎟
u

⎝ n−b ⎠ ⎝ n−b ⎠
a
(n − b )β
zm
n −b

1− m
z= [Re β > 0, Re n − b > 0]
n−b

—Sect. 3.462—
 136

⎛ a ⎞ Integral follows from

∞ n+b Γ⎜ z m ⎟ ⎛ a ⎞ ⎛ a ⎞
e − βx ⎝ n+b⎠ v
e − βx
n+b Γ⎜ z m
dx = ⎝
n+b
, β u n + b ⎟ − Γ⎜ z m
⎠ ⎝ n+b
, βv n + b ⎟
⎠, 1− m

∫0 x m± a dx = ∫ m± a
z=
u x
a
n+b
(n + b)β n + b
zm
a
(n + b )β
zm [Re β > 0, Re n + b > 0]
n +b

1− m ⎡ a ⎤ by substituting,
z= ⎢Re β > 0, Re n + b > 0, Re z m n + b > 0⎥
n+b ⎣ ⎦ u = 0, v = ∞, Γ(u , β u n + b ) =
Γ(u,0) = Γ(u ), Γ(v, β v n + b ) =
Γ(v, ∞ ) = 0.
⎛ a ⎞ Integral follows from
∞ − βx n − b
Γ⎜ z m ⎟
e ⎝ n−b⎠
∫
0 x
m± a
dx = a v

∫
e − βx
n−b
⎛
Γ⎜ z m
dx = ⎝
a
n−b
⎞
⎠
⎛
, β u n − b ⎟ − Γ⎜ z m
⎝
a
n−b
⎞
, βv n −b ⎟
⎠, 1− m

(n − b )β
zm z=
n −b m±a
n−b
a
u x
(n − b )β n − b
zm

[Re β > 0, Re n − b > 0]

1− m ⎡ a ⎤
⎢Re β > 0, Re n − b > 0, Re z m n − b > 0⎥
z= by substituting,
n−b ⎣ ⎦ u = 0, v = ∞, Γ(u , β u n + b ) =
Γ(u,0) = Γ(u ), Γ(v, β v n + b ) =
Γ(v, ∞ ) = 0.
⎛ a ⎞ Integral follows from
∞ n+b Γ⎜ z m , β u n +b ⎟
e − βx n+b ⎠
∫u x m± a dx = ⎝ a v
e − βx
n+b
⎛
Γ⎜ z m
a
n+b
⎞ ⎛
, β u n − b ⎟ − Γ⎜ z m
a
n+b
⎞
, βv n + b ⎟
⎝ ⎠ ⎝ ⎠,
(n + b )β ∫u x m ± a
zm dx =
n +b a

(n + b )β n −b
zm

1− m
[Re β > 0, Re n + b > 0]
1− m z=

z= [Re β > 0, Re n + b > 0] n+b

by substituting,
n+b

(
u = u , v = ∞, Γ u , β u n + b = Γ u , β u n + b , ) ( )
( )
Γ v, β v n + b = Γ(v, ∞ ) = 0.
⎛ a ⎞ Integral follows from
∞ − βx n − b
Γ⎜ z m , β u n −b ⎟
n−b
dx = ⎝ ⎠
e
∫ x m± a a v
e − βx
n−b
⎛
Γ⎜ z m
⎝
a
n−b
⎞
⎠
⎛
, β u n − b ⎟ − Γ⎜ z m
⎝
a
n−b
⎞
, βv n +b ⎟
⎠,
(n − b )β ∫u x m ± a
zm dx =
u n −b a

(n − b)β n −b
zm

1− m
, [Re β > 0, Re n − b > 0]
1− m z=

z= [Re β > 0, Re n − b > 0] n−b

by substituting,
n−b

(
u = u , v = ∞, Γ u , β u n − b = Γ u , β u n − b , ) ( )
( )
Γ v, βv n−b = Γ(v, ∞ ) = 0.

—Sect. 3.462—
 137

⎛ a ⎞ Integral follows from

n+b γ⎜z m , βu n+b ⎟ ⎛ a ⎞ ⎛ a ⎞
e − βx
, βu n + b ⎟ − Γ⎜ z m , βv n + b ⎟
1− m
Γ⎜ z m
n+b
u
dx = ⎝ ⎠
n+b
v
e − βx n+b n+b
dx = ⎝ ⎠ ⎝ ⎠

∫0 x m± a a
z=
n+b
∫
u x
m±a

(n + b )β n + b
zm
a

(n + b )β
zm 1− m
n +b z= , [Re β > 0, Re n + b > 0]
n+b

⎡ a ⎤
⎢Re β > 0, Re n + b > 0, Re z m n + b > 0⎥ by substituting, u = 0, v = u for the
⎣ ⎦
quantities
resulting in the observed lower
incomplete
gamma function, γ (a, x ), derived from,
(
Γ zm
a
n+b
) (
,0 −Γ z m
a
n+b
, βu n + b )
a
zm
n +b β n +b
( )

⎛ a ⎞ Integral follows from

− βx n − b
γ⎜z m , β u n −b ⎟ ⎛ a ⎞ ⎛ a ⎞
n−b 1− m Γ⎜ z m , β u n + b ⎟ − Γ⎜ z m , βv n −b ⎟
u
dx = ⎝ ⎠
e n −b

∫0 x m±a
e − βx ⎝ n−b ⎠ ⎝ n−b ⎠ 1− m
z=
v

∫u x m ± a dx = a
z=
n−b
n−b (n − b )β
zm
a n −b

(n − b )β
zm
n −b [Re β > 0, Re n − b > 0]

⎡ a ⎤
⎢⎣Re β > 0, Re n − b > 0, Re z m n − b > 0⎥⎦ by substituting, u = 0, v = u for the
quantities resulting in the observed
lower incomplete gamma function,
γ (a, x ), derived from,
Γ zm ( a
n −b
)(
,0 −Γ z m
a
n −b
, βu n −b )
a
zm
( n − b ) β n −b

—Sect. 3.462—
 138

v
Extension of integrals 3.462.12 –
∫ (ax ± b ) e dx =
m − ρx
3.462.14 with same integrand.
u
• Apply three successive change of
ρb
± variable substitutions:
a e m
⎡ ⎛ a
ρb ⎞ ⎛ ρb ⎞⎤
⎢ Γ⎜ m + 1, ρu ± ⎟ − Γ⎜ m + 1, ρv ± ⎟
a ⎠⎥⎦
t
ρ m +1
⎣ ⎝ a ⎠ ⎝ s = ρx; t = as ± bρ ; w = ,
a
⎡ ⎛b ⎞ ⎛b ⎞ ⎤ to obtain,
⎢ ρ > 0, arg⎜ ± u ⎟ < π , arg⎜ ± v ⎟ < π ⎥ ρb ⎡ ρv ± ρb
⎣ ⎝a ⎠ ⎝a ⎠ ⎦ ± ⎤
a e a ⎢ a m −w ⎥
m

ρ m +1 ⎢ ∫ρb
I= w e dw⎥
⎢ ρu ± ⎥
⎣ a ⎦
• Use same indefinite-integral
approach outlined in Sect. 3.323 for
derivation of
v
( )dx ,
∫e
− ax 2 + kbx + c

u
ρb
± ρb
m ρv ±
[− Γ(m + 1, w)]ρu ± ρab
a
a e
giving, I =
ρ m +1 a
which leads to integral.
o Easier derivation: set
ds
s = ax ± b, = a,
dx
su = au ± b, sv = av ± b
to obtain,
ρb
ρs
e± a
sv
−
I= ∫s e
m a
ds ,
a su

and proceed as above.

¾ NOTE: see application of integral
in Sect. 3.471, for solution of:
v
(a − x )μ −1 exp⎛ − β ⎞dx
∫u x μ +1 ⎜
⎝ x⎠
⎟

Solution requires two changes of

variables, the second of which is
v
provided by ∫ (ax ± b ) e − ρx dx .
m

—Sect. 3.462—
 139

v
Derivation similar to
∫ (b − ax ) e dx =
m − ρx
v

∫ (ax ± b ) e − ρx dx =
m
u
u
ρb ρb
− ±

(− 1)m a em+1 ⎡⎢Γ⎛⎜ m + 1, ρu − ρb ⎞⎟ − Γ⎛⎜ m + 1, ρv − ρb ⎞⎟⎤⎥

m a a me a
⎡ ⎛ ρb ⎞ ⎛ ρb ⎞⎤
ρ m +1 ⎢Γ⎜ m + 1, ρu ± a ⎟ − Γ⎜ m + 1, ρv ± a ⎟⎥
⎣ ⎝ ⎠ ⎝ ⎠⎦
ρ ⎣ ⎝ a ⎠ ⎝ a ⎠⎦
• Apply three successive change of
⎡ ⎛ b⎞ ⎛ b⎞ ⎤ variable substitutions:
⎢ ρ > 0, arg⎜ u − ⎟ < π , arg⎜ v − ⎟ < π ⎥
⎣ ⎝ a⎠ ⎝ a⎠ ⎦ s = ρx; t = bρ − as; w = ,
t
a
to obtain,
ρb ⎡ ρb − ρv ⎤
±
a e a ⎢a
m
⎥
I = m +1 ⎢ ∫ wm e w dw⎥
ρ
⎢ ρb − ρu ⎥
⎣a ⎦
• Use indefinite-integral approach
previously submitted,
(− 1)1−γ Γ(γ ,−ax n ) γ = m + 1
∫ =
n
m ax
x e dx γ
na n
which leads to integral.
Alternatively, from above integral,
v

∫ (ax ± b )m e − ρx dx ,
u
v v
.
∫ (b − ax ) e − ρx dx =(− 1) ∫ (ax − b ) e − ρx dx
m m m

u u

Easier derivation: set

ds
s = b − ax, = − a,
dx
su = b − au , sv = b − av
to obtain,
ρb
− sv ρs
a
e
a su∫
I =− s e a ds , m

and proceed as above.

∞ −
ρb Set u = 0, v = ∞ in above integral,
m a
⎛ ρb ⎞
∫ (b − ax ) dx = (− 1)
− ρx a e
Γ⎜ m + 1, −
m m v
e ⎟ ∫ (b − ax )
m
e − ρx dx =
0
ρ m +1
⎝ a ⎠ u
ρb
−
⎡ ⎛b⎞ ⎤ ⎡ ⎛ ρb ⎞ ⎛ ρb ⎞ ⎤
m a
(− 1)m a e
⎢Γ⎜ m + 1, ρu − a ⎟ − Γ⎜ m + 1, ρv − a ⎟⎥
⎢ ρ > 0, arg⎜ ⎟ < π ⎥ ρ m +1
⎣ ⎝ ⎠ ⎝ ⎠⎦
⎣ ⎝a⎠ ⎦

—Sect. 3.462—
 140

u Set u = 0, v = u in above integral,

∫ (b − ax ) e dx =
m − ρx
v

∫ (b − ax ) e − ρx dx =
m
0
u
ρb ρb
− −
⎡ ⎛ ρb ⎞ ⎛ ρb ⎞ ⎤
⎡ ⎛ ρb ⎞ ⎛ ρb ⎞ ⎤ a me a
(− 1)
m a
Γ⎜ m + 1, ρu − ⎟ − Γ⎜ m + 1, ρv −
(− 1) a e ⎟
m

Γ⎜ m + 1,− ⎟ − Γ⎜ m + 1, ρu − ⎟ ρ m +1 ⎢⎣ ⎝ a ⎠⎥⎦
m
a ⎠ ⎝
m +1 ⎢
ρ ⎣ ⎝ a ⎠ ⎝ a ⎠⎥⎦
⎡ ⎛ b⎞ ⎤
⎢u > 0, ρ > 0, arg⎜ u − ⎟ < π ⎥
⎣ ⎝ a⎠ ⎦

∞ −
ρb Set v = ∞ in above integral,
a me ⎡ ⎛ pb ⎞⎤
∫ (b − ax ) e − ρx dx = (− 1)
a v

⎢Γ⎜ m + 1, pu − a ⎟⎥ ∫ (b − ax )
m m
e − ρx dx =
m

u
ρ m+1 ⎣ ⎝ ⎠⎦ u
ρb
−
⎡ ⎛ ρb ⎞ ⎛ ρb ⎞ ⎤
⎡ ⎤
m a

⎛ b⎞ (− 1)m a e
⎢Γ⎜ m + 1, ρu − a ⎟ − Γ⎜ m + 1, ρv − a ⎟⎥
⎢ ρ > 0, arg⎜ u − ⎟ < π ⎥ ρ m +1
⎣ ⎝ ⎠ ⎝ ⎠⎦
⎣ ⎝ a⎠ ⎦

e − ρx
v
Extension of integrals 3.462.15 –
∫ (ax ± b ) n
dx = 3.462.17 with same integrand.
u
• Same approach as preceding
ρb
± integral,
ρ n −1
e ⎡ ⎛a
ρb ⎞ ⎛ ρb ⎞ ⎤
⎢Γ⎜ − n + 1, ρu ± a ⎟ − Γ⎜ − n + 1, ρv ± a ⎟⎥
v

∫ (ax ± b ) e − ρx dx =
m

an ⎣ ⎝ ⎠ ⎝ ⎠⎦ u
ρb
⎡ ⎛b ⎞ ⎛b ⎞ ⎤ a me
±
a
⎡ ⎛ ρb ⎞ ⎛ ρb ⎞ ⎤
⎢ρ > 0, arg⎜ ± u ⎟ < π , arg⎜ ± v ⎟ < π ⎥ Γ⎜ m + 1, ρu ±
ρ m +1 ⎢⎣ ⎝ a ⎠
⎟ − Γ⎜ m + 1, ρv ±
⎝
⎟
a ⎠⎥⎦
⎣ ⎝a ⎠ ⎝a ⎠ ⎦
resulting in,
±
ρb ⎡ ρv ± ρab ⎤
ρ n −1e a
⎢ ⎥
⎢ ∫ w e dw⎥ ⇒
−n − w
I=
an
⎢⎣ ρu ± ρab ⎥⎦
ρb
± ρb
ρ e n −1
[− Γ(− n + 1, w)]ρu ± ρab ,
a ρv ±
I=
an a
which leads to integral.

NOTE: see formula in Sect. 3.362

below for algebraic function integral,
1
with n = .
2
e − ρx Derivation similar to
v

∫ (b − ax ) n
dx = v
(b − ax )m e − ρx dx =
ρb ⎡ ⎛ ρb ⎞ ⎛ ρb ⎞⎤
u
∫ ame
−
a ⎢Γ⎜ m + 1, ρu − a ⎟ − Γ⎜ m + 1, ρv − a ⎟⎥
−
ρb u (− 1) m ⎣ ⎝ ⎠ ⎝ ⎠⎦
ρ m +1
ρ n −1 e a
⎡ ⎛ ρb ⎞ ⎛ ρb ⎞⎤
(− 1) n
⎢Γ⎜ − n + 1, ρu − a ⎟ − Γ⎜ − n + 1, ρv − a ⎟⎥
an ⎣ ⎝ ⎠ ⎝ ⎠⎦ Alternatively, from above integral,
⎡ ⎤ e − ρx , v e − ρx dx =(− 1)n v e − ρx dx .
v
⎛ b⎞ ⎛ b⎞ ∫ (ax ± b ) ∫ (b − ax ) ∫ (ax − b)
⎢ ρ > 0, arg⎜ u − ⎟ < π , arg⎜ v − ⎟ < π ⎥
dx n n n

⎝ a⎠ ⎝ a⎠
u u u

⎣ ⎦

—Sect. 3.462—
 141

∞ −
ρb Set u = 0, v = ∞ in above integral,
e − ρx n ρ
n −1 a
⎛ ρb ⎞
∫0 (b − ax )n dx = (− 1) e
Γ⎜ − n + 1, − ⎟
v

∫(
e − ρx
dx =
an ⎝ a ⎠ u
b − ax )n

ρb
v −
e − ρx ρ n −1e
⎡ ⎤
a ⎡ ⎛ ρb ⎞⎤
∫(
ρb ⎞ ⎛
⎛b⎞ b − ax )n
dx = (− 1)n ⎢Γ⎜⎜ − n + 1, ρu − ⎟ − Γ⎜⎜ − n + 1, ρv −
a ⎟⎠
⎟⎥
a ⎟⎠⎥⎦

⎢ ρ > 0, arg⎜ ⎟ < π ⎥

an ⎣⎢ ⎝ ⎝
u

⎣ ⎝a⎠ ⎦
u
e − ρx Set u = 0, v = u in above integral,
∫0 (b − ax )n dx = v

∫(
e − ρx
dx =
b − ax )n
u
ρb ρb
− v
e − ρx ρ n −1e
−

ρ n −1e ρb ⎞ ⎛ ρb ⎞⎤
a ⎡ ⎛ ρb ⎞⎤
⎡ ⎛ ∫(
ρb ⎞ ⎛
a dx = (− 1)n ⎢Γ⎜⎜ − n + 1, ρu − ⎟ − Γ⎜⎜ − n + 1, ρv − ⎟⎥

(− 1)n ⎢ Γ⎜ − n + 1,− ⎟ − Γ⎜ − n + 1, ρu − ⎟
b − ax )n an ⎢⎣ ⎝ a ⎟⎠ ⎝ a ⎟⎠⎥⎦

a ⎠⎥⎦
u

⎣ ⎝ a ⎠ ⎝
n
a
⎡ ⎛ b⎞ ⎤
⎢u > 0, ρ > 0, arg⎜ u − ⎟ < π ⎥
⎣ ⎝ a⎠ ⎦
∞ −
ρb Set v = ∞ in above integral,
e − ρx ρ n −1e a ⎡ ⎛
∫
ρb ⎞ ⎤
dx = (− 1)n v
⎢Γ⎜⎜ − n + 1, ρu − ⎟⎥ e − ρx
(b − ax )n a n ⎢
⎣ ⎝ a ⎟⎠⎥⎦ ∫( b − ax )n
dx =

u u
ρb
v −
⎡ ⎛ b⎞ ⎤ ∫(
e − ρx
dx = (− 1)n
ρ n −1e a ⎡ ⎛
⎢Γ⎜⎜ − n + 1, ρu −
ρb ⎞ ⎛
⎟ − Γ⎜⎜ − n + 1, ρv −
ρb ⎞⎤
⎟⎥

⎢ ρ > 0, arg⎜ u − ⎟ < π ⎥

b − ax )n an ⎣⎢ ⎝ a ⎟⎠ ⎝ a ⎟⎠⎥⎦
u

⎣ ⎝ a⎠ ⎦
n+q
v ⎛ x−a ⎞
m ± p − β ⎜⎜ ⎟⎟
Solve integral separately for
∫
⎛x−a⎞
⎜⎜ ⎟⎟
⎝ b ⎠
e ⎝ b ⎠ dx = x m + p & x m − p and assign signs for general
u
⎡ n+q ⎤ ⎡ n+q ⎤ expression.
p ⎛u −a⎞ ⎥ − Γ ⎢γ ± p , β ⎛⎜ v − a ⎞⎟
Γ ⎢γ ± , β⎜ ⎟ ⎥
⎢ n + q ⎜⎝ b ⎟⎠ ⎥ ⎢ n + q ⎜⎝ b ⎟⎠ ⎥
b ⎣ ⎦ ⎣ ⎦
γ±
p Set
x − a ds 1 u−a
(n + q )β n + q s= , = , su = ,
b dx b b
m +1 v−a
to obtain,
γ = sv = ,
n+q b
⎡− ∞ < Re a < +∞, Re b > 0, ⎤ sv
⎢ ⎥
I = b ∫ s m ± p e − βs ds. Now use formula
n+q
⎢ ⎛u −a⎞ ⎛v−a⎞ ⎥
⎢ arg⎜⎜ b ⎟⎟ < π , arg⎜⎜ b ⎟⎟ < π ,⎥
⎢ ⎝ ⎠ ⎝ ⎠ ⎥
⎢Re β > 0, Re n + q > 0 ⎥ su
⎣ ⎦
v

∫x
n+b
above, this section, m±a
e − βx dx
u

with
⎡ m + 1⎤
⎢⎣m = m, a = p, n = n, b = q, γ = n + b ⎥⎦
¾ Not verifiable on math.com
m± p ⎛ x−a ⎞
n−q
Same as ⎛ x − a ⎞ with n − q. m± p ⎛ x−a ⎞
n+q

⎛ x−a⎞ −β ⎜
v −β ⎜ ⎟
⎟
v ⎝ b ⎠
dx = ∫u ⎜⎝ b ⎟⎠ e ⎝ b ⎠
dx
⎜ ⎟ e
∫⎝ b ⎠
u
¾ Not verifiable on math.com
⎡ p ⎛u −a⎞
n−q
⎤ ⎡ p ⎛v−a⎞
n−q
⎤
Γ ⎢γ ± ,β⎜ ⎟ ⎥ − Γ ⎢γ ± ,β⎜ ⎟ ⎥
⎢⎣ n−q ⎝ b ⎠ ⎥⎦ ⎢⎣ n−q ⎝ b ⎠ ⎥⎦
b p
γ±
(n − q )β n−q

—Sect. 3.462—
 142

m +1
γ =
n−q
⎡ ⎛u−a ⎞ ⎤
⎢− ∞ < Re a < +∞, Re b > 0, arg⎜⎜ ⎟⎟ < π ,⎥
⎢ ⎝ b ⎠ ⎥
⎢ ⎥
⎢ ⎛v−a⎞ ⎥
⎢ arg⎜⎜ b ⎟⎟ < π , ⎥
⎢ ⎝ ⎠ ⎥
⎢Re β > 0, Re n − q > 0 ⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦

⎡ ⎛⎜ x − a ⎞⎟ n ⎤ ⎛ x − a ⎞ ds 1
⎛ x−a⎞
m
∞ Set s = ⎜ ⎟, = , s0 = -a/b and substitute.
∫ 0
⎜
⎝ b ⎠
⎟ exp ⎢− β ⎜⎜
⎢ ⎝ b ⎟⎠ ⎥
⎟ ⎥ dx = ⎝ b ⎠ dx b
⎣ ⎦ dt
Then,set t = βs n , = nβs n−1 ,
bΓ⎛⎜ γ , β (− a / b )n ⎞⎟
ds
⎝ ⎠ 1 n,
⎛t⎞
nβ γ s = ⎜⎜ ⎟⎟ , t −a / b = β (− a / b) ,
n

γ =
m +1 ⎝β ⎠
n
m +1
[re β > 0, re n > 0, re b > 0,-∞ < re a < +∞, γ= , and substitute.
arg (− a / b ) < π ]. n
n+ q
⎛ x−a ⎞
v − β ⎜⎜ ⎟⎟
e ⎝ b ⎠
∫ ⎛ x−a ⎞
⎜⎜ ⎟⎟
m± p
dx = Solve above integral,
⎜x − a⎟
⎛ ⎞ n+q
u
⎝ b ⎠ v ⎛ x − a ⎞ m ± p − β ⎜⎜ b ⎟⎟ , with
⎝ ⎠
∫⎜ ⎟ e dx
⎡ n+ q ⎤ ⎡ n+ q ⎤
p ⎛u−a⎞ ⎥ − Γ ⎢ z m p , β ⎛⎜ v − a ⎞⎟ u⎝ b ⎠
Γ⎢ z m , β ⎜⎜ ⎥
m ± p set to − (m ± p ).
⎟⎟ ⎜ ⎟
⎢ n+q ⎝ b ⎠ ⎥ ⎢ n+q ⎝ b ⎠ ⎥
b ⎣ ⎦ ⎣ ⎦
p
zm
(n + q )β n + q
1− m
¾ Not verifiable on math.com
z=
n+q
⎡ ⎛u−a⎞ ⎤
⎢ − ∞ < Re a < +∞, Re b > 0, arg⎜⎜ ⎟⎟ < π ,⎥
⎢ ⎝ b ⎠ ⎥
⎢ ⎥
⎢ ⎛v−a⎞ ⎥
arg⎜
⎢ ⎜ b ⎟ ⎟ < π , ⎥
⎢ ⎝ ⎠ ⎥
⎢Re β > 0, Re n + q > 0 ⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
n+q
n−q ⎛ x−a ⎞
⎛ x−a ⎞ −β ⎜ ⎟
−β ⎜ ⎟ Similar to above integral, v
e ⎝ b ⎠
v
e ⎝ b ⎠ ∫ ( ) dx

∫ ( )
m± p
x−a

m± p
dx = u b
x−a
u b
¾ Not verifiable on math.com
⎡ p ⎛u −a⎞ ⎤
n−q
⎡ p ⎛v−a⎞ ⎤
n−q

Γ⎢ z m ,β⎜ ⎟ ⎥ − Γ ⎢ z m , β ⎜ ⎟ ⎥
⎢⎣ n − q ⎝ b ⎠ ⎥⎦ ⎢⎣ n − q ⎝ b ⎠ ⎥⎦
b p
zm
(n − q )β n−q

1− m
z=
n−q
⎡ ⎛u −a⎞ ⎤
⎢− ∞ < Re a < +∞, Re b > 0, arg⎜ ⎟ < π ,⎥
⎢ ⎝ b ⎠ ⎥
⎢ ⎛v−a⎞ ⎥
⎢ arg⎜ ⎟ < π , ⎥
⎢ ⎝ b ⎠ ⎥
⎢Re β > 0, Re n − q > 0 ⎥
⎢ ⎥
⎢⎣ ⎥⎦

—Sect. 3.462—
 143

Formula 3.462.10 In integral, set

∞
(1 − aβ ) s = ( x − a),
ds
= 1, s 0 = −a, s ∞ = ∞;
∫ ( x − a )e
− β ( x −a )
dx = e aβ
0
β2 dx
to get
[Re β > 0] ∞ ∞
1 1
I = ∫ se ds ⇒ 2 ∫ β se − βs ds.
− βs

β −a β −a
dt
Set t = β s,
ds
= β , t − a = − β a , t ∞ = ∞, giving the
result,
∞
te − t dt = Γ(2,−aβ )
∫β
1
I= since
β2 β2
−a

Γ(n, x ) =
n −1
xm
(n − 1)! e − x ∑ , [n = 1,2,...]
m =0 m!

⇒
e aβ
(1 − aβ ) .
β2
Formula 3.462.11 ∞
See above, ∫ ( x − a)e −β ( x−a ) dx for
∞
(1 − aβ )
∫ ( x − a )e −β ( x+ a ) dx = e −aβ
β2
0
0 similar proof
[Re β > 0]
Formula 3.462.12 In above integral,
pb v
±
∫ (ax ± b )
∞
e − ρx dx =
m
a me ⎛ a
pb ⎞
∫ (ax ± b ) e − px dx = Γ⎜ m + 1,± ⎟
m
m +1 u
0 p ⎝ a ⎠ ρb
m ±
⎡ b ⎤ a e a ⎡ ⎛ ρb ⎞ ⎛ ρb ⎞⎤
⎢ p > 0, arg a < π ⎥ m+1 ⎢Γ⎜ m + 1, ρu ± a ⎟ − Γ⎜ m + 1, ρv ± a ⎟⎥ ,
ρ ⎣ ⎝ ⎠ ⎝ ⎠⎦
⎣ ⎦
set u = 0, v → ∞.
Formula 3.462.13 Proof similar to above for integral,
±
pb ∞

∫ (ax ± b)
∞
⎛ pb ⎞ e − px dx
m a m

∫ (ax ± b )
− px a e
dx = Γ⎜ m + 1, pu ±
m
e m +1 ⎟
u
p ⎝ a ⎠ 0

⎡ b ⎤
⎢ p > 0, arg a ± u < π ⎥
⎣ ⎦
Proof similar to above for integral,
Formula 3.462.14 ∞

∫ (ax ± b) e − px dx
m
u

∫ (ax ± b )
0
e − px dx =
m

pb
±
a e m
⎡ ⎛ a
pb ⎞ ⎛ pb ⎞⎤
Γ m + 1,±
m +1 ⎢ ⎜ ⎟ − Γ⎜ m + 1, pu ± ⎟
p ⎣ ⎝ a ⎠ ⎝ a ⎠⎥⎦

—Sect. 3.462—
 144

⎡ b ⎤
⎢u > 0, p > 0, arg a ± u < π ⎥
⎣ ⎦
Formula 3.462.15 In above integral,
pb
e − ρx
v
±
∞
e − px p e n −1 a
⎛ pb ⎞ ∫ (ax ± b ) dx =
∫0 (ax ± b)n dx = a n Γ⎜ − n + 1,±
n
⎟ u

⎝ a ⎠ ρb
n −1 ± a
ρ e ⎡ ⎛ ρb ⎞ ⎛ ρb ⎞ ⎤
⎡ b ⎤ n ⎢Γ⎜ − n + 1, ρu ± ⎟ − Γ⎜ − n + 1, ρv ± ⎟⎥ ,
⎣ ⎝ ⎠ ⎝ ⎠⎦
⎢ p > 0, arg a < π ⎥
a a a

⎣ ⎦ set u = 0, v → ∞.
Formula 3.462.16 Proof similar to above for integral,
pb ∞
∞ − px n −1
±
e − px
⎛ pb ⎞ ∫0 (ax ± b)n dx
a
e p e
∫ (ax ± b )
u
n
dx =
an
Γ⎜ − n + 1, pu ±
⎝ a ⎠
⎟

⎡ b ⎤
⎢ p > 0, arg a ± u < π ⎥
⎣ ⎦
Formula 3.462.17 Proof similar to above
e − px ∞
u
e − px
∫0 (ax ± b )n dx = integral, ∫ dx
0 (ax ± b )
n

pb
±
n −1
p e ⎡ ⎛ a
pb ⎞ ⎛ pb ⎞⎤
⎢Γ⎜ − n + 1,± a ⎟ − Γ⎜ − n + 1, pu ± a ⎟⎥
an ⎣ ⎝ ⎠ ⎝ ⎠⎦
⎡ b ⎤
⎢u > 0, p > 0, arg a ± u < π ⎥
⎣ ⎦
∞
(1 + aβ ) ∞

∫ (x + a )e
−β ( x−a )
dx = e aβ
See above, ∫ ( x − a)e −β ( x−a ) dx for
0 β2 0
[Re β > 0] similar proof
⎛ p ⎛ a⎞
n+ q
⎞ ⎛ x − a ⎞ ds 1
bΓ⎜ γ + ,β⎜− ⎟ ⎟ Set s = ⎜ ⎟, = and substitute.
∞ m+ p ⎛ x−a ⎞
n+ q
⎜ n+q ⎝ b⎠ ⎟ ⎝ b ⎠ dx b
⎛x−a⎞ −β ⎜
dx = ⎝ ⎠
⎟

∫0 ⎜⎝ b ⎟⎠ e ⎝ b ⎠
γ+
p Set t = β s n + q ,
(n + q )β n+q
dt ⎛ t
= (n + q )βs n + q −1 , s = ⎜⎜
⎞
⎟⎟
1 /(n + q)
,
⎝β
m +1 ⎡− ∞ < Re a < +∞, Re b > 0, ⎤ ds ⎠
γ = ⎢ ⎥ γ=
m +1
n+q ⎣ arg(− a / b) < π , Re β > 0, Re n + q > 0⎦
n+q
∞ γ + p −1
b
∫t
n+q −t
to get, p
e dt.
γ+
(n + q )β n+q 0

Apply definition of upper incomplete

Gamma function, Γ(a, x ), Form. 8.350.1

⎛ p ⎛ a ⎞ ⎞⎟
n+q See above solution
n+q ⎜
bΓ γ − ,β⎜− ⎟
∞ m− p − β ⎛ x −a ⎞ ⎜ n + q ⎝ b ⎠ ⎟⎠
⎛x−a⎞ ⎜ ⎟
⎝
∫ ⎜
⎝ b
⎟ e
⎠
⎝ b ⎠
dx =
γ+
p
n+q
(n + q) β
0

—Sect. 3.462—
 145

m +1
γ = [−∞ < Re a < +∞, Re b > 0,
n+q
arg(− a / b) < π , Re β > 0, Re n + q > 0]

Formula 3.462.18 Similar to above integrals

∞
⎛x−a⎞ ⎛ x−a⎞
j k

∫0 ⎜⎝ b ⎟⎠ exp β ⎜⎝ b ⎟⎠ dx =
⎡ j + 1 ⎛ a ⎞k ⎤
Γ⎢ , β⎜− ⎟ ⎥
⎣⎢ k ⎝ b ⎠ ⎦⎥
b j +1
kβ k

⎡ ⎡ ⎛ a ⎞⎤ ⎤
⎢Re b > 0, Re β > 0, Re k > 0, ⎢arg⎜ − ⎟⎥ > 0⎥
⎣ ⎣ ⎝ b ⎠⎦ ⎦

⎛ ⎛ u − a ⎞ ⎞⎟
n+q ⎛ x − a ⎞ ds 1
p Set s = ⎜⎜ ⎟⎟; = and substitute.
⎜
bΓ γ + ,β⎜ ⎟ ⎟ ⎝ b ⎠ dx b
⎜
n+q
∞
⎛x−a⎞
m+ p ⎛ x −a ⎞
−β ⎜ ⎟
⎝ n + q ⎝ b ⎠ ⎠ Set t = βs n + q ;

∫u ⎜⎝ b ⎟⎠ e ⎝ ⎠ dx =
b
γ+
p dt
(n + q )βs n + q −1,
(n + q )β n−q
ds
1 /(n + q)
⎛ t ⎞ m +1
s = ⎜⎜ ⎟⎟ ,γ =
m +1
to get
⎝β⎠ n+q
γ = [−∞ < Re a < +∞, ∞
n+q
p
γ+ −1

∫ + q e −t dt.
b n
t
p
⎛u −a⎞ γ+
( )
n+q β u −a n+q
Re b > 0, arg⎜ ⎟ < π , Re β > 0, Re n + q > 0] ( n + q) β b

⎝ b ⎠ Apply definition of upper incomplete gamma function,

Γ(α , x) of
Formula 8.350 2.

⎛ ⎛ u − a ⎞ ⎞⎟
n−q See Solution above
p
bΓ⎜ γ + ,β⎜ ⎟
⎜ n − q ⎝ b ⎠ ⎟⎠
n−q
∞ m+ p ⎛ x −a ⎞
⎛x−a⎞ −β ⎜
dx = ⎝
⎟

∫u ⎜⎝ b ⎟⎠ e ⎝ b ⎠
γ+
p

(n − q )β n −q

m +1
γ = [−∞ < Re a < +∞, Re b > 0,
n−q
⎛u −a⎞
arg⎜ ⎟ < π , Re β > 0, Re n − q > 0]
⎝ b ⎠
⎛ p ⎛ a ⎞ ⎞⎟
n−q See above solution
n−q ⎜
bΓ γ + , β⎜− ⎟
∞ m+ p − β ⎛ x −a ⎞ ⎜ n − q ⎝ b ⎠ ⎟⎠
⎛x−a⎞ ⎜ ⎟
⎝
∫ ⎜
⎝ b
⎟ e
⎠
⎝ b ⎠
dx =
γ+
p
n −q
(n − q)β
0

m +1
γ = [−∞ < Re a < +∞, Re b > 0,
n−q
arg(−a / b) < π , Re β > 0, Re n − q > 0]

—Sect. 3.462—
 146

⎛ p ⎛ a⎞ ⎞
n−q See above solution
n−q bΓ ⎜ γ − , β⎜− ⎟ ⎟
∞ m− p − β ⎛⎜ x − a ⎞⎟ ⎜ n − q ⎝ b ⎠ ⎟⎠
⎛x−a⎞
∫⎜ ⎟ e ⎝ b ⎠
dx = ⎝
0⎝
b ⎠ p
γ−
n −q
(n − q) β
m +1
γ = [−∞ < Re a < +∞, Re b > 0,
n−q
arg(−a / b) < π , Re β > 0, Re n − q > 0]
Γ( z ) In Formula 3.326.2, set m → −m.
∞
e − βx
n

∫
0
x m
dx =
nβ z
1− m
z= [ Re β > 0 , Re n > 0, Re z > 0]
n
u
e − βx
n
(
γ z, βu n ) In Sect. 3.326, Formula
∫ xm
dx =
nβ z
v
exp⎛⎜ − β x n ⎞⎟
0

1− m
∫ ⎝
xm
⎠ dx =

z= [u > 0, Re β > 0, Re n > 0, Re z > 0]

u

n Γ⎛⎜ z , βu n ⎞⎟ − Γ⎛⎜ z , βv n ⎞⎟
⎝ ⎠ ⎝ ⎠ 1− m
z=
nβ z n

set u → 0, v → v.

Formula 3.462.19
Γ(z , βu n )
∞ − βx n In Sect. 3.326, Formula
e
∫u x m dx = v
(
exp − βx n )
Γ z, βu n − Γ z, βv n ( ) ( )
nβ z ∫u x m dx = nβ z
1− m
z= [Re β > 0, Re n > 0] 1− m
n z=
n

set u → u, v → ∞.

—Sect. 3.462—
 147

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
∞
⎛ − μx 2 − βx 2 ⎞ dx Using Formula 3.461 5,
∫ ⎜e −e ⎟ 2=
u⎝ ⎠x ⎛ 1 ⎞
Γ⎜ − , μ u 2 ⎟
− μu 2
− βv 2 ∞
⎝ 2 ⎠ from derived formula,
[ ( )] [ ( )]
−e − μx
2 dx
e ∫e =
+ βπ 1− Φ u β − μπ 1− Φ u μ u x2 −
1
u 2μ 2

[Re μ > 0, Re β > 0] ∞

e − βx n
(
Γ z, βu n )
∫
u
x m
dx =
nβ z
1− m
z= [Re β > 0, Re n > 0]
n
⎡ 1⎤
with parameters ⎢m = 2, β = μ , n = 2, z = − ⎥ .
⎣ 2⎦
Using formula in Sect. 8.359, below,
⎛ 1 ⎞ ⎡ e−x ⎛ 1 ⎞⎤
Γ⎜ − , x ⎟ = 2 ⎢ − Γ⎜ , x ⎟ ⎥ ,
⎝ 2 ⎠ ⎣ x ⎝ 2 ⎠⎦
⎛ 1 ⎞
Γ⎜ − , μ u 2 ⎟
so that ⎝ ⎠ = e− μu
2
2
1 − μπ [1 − Φ(u μ )],
− u
2μ 2
and likewise, subtracting off a second term,
⎛ 1 ⎞
Γ⎜ − , β u 2 ⎟ − βu 2
∞
∫e
− βx
2 dx
= ⎝ 2 ⎠ =e
u
[ ( )]
− βπ 1 − Φ u β ,
u x2 −
1
2β 2

so that,
∞
⎛ − μx 2 − βx 2 ⎞ dx
∫ ⎜e −e ⎟ 2=
u⎝ ⎠x
− μu 2 − βv 2
e −e
u
[ ( )]
+ πβ 1 − Φ u β − πμ 1 − Φ u μ . [ ( )]
NOTE: As a check, the calculation, for
u = 0, Φ (0 ) = 0, yields,
∞
⎛ − μx 2 − βx 2 ⎞ dx
∫ ⎜e
0⎝
−e ⎟ 2 = πβ − πμ = π
⎠x
( β− μ, )
Formula 3.464

¾ Verified on math.com
− βu
2
− μu
2
e − μx
2
e − βx
2
−e
u
u
⎛ − μx 2 − βx 2 ⎞ dx e Solve u
separately.
∫ ⎜e −e ⎟ 2= + ∫ 2 dx & ∫ 2 dx
0⎝ ⎠x u 0 x x
[ ( )] [ ( )]
o

βπ Φ u β − μπ Φ u μ [u > 0, Re μ > 0, Re β > 0] Using indefinite integral previously submitted,

——Sect. 3.464——
 148

β z Γ⎛⎜ − z, βx n ⎞⎟
e − βx
n

∫ dx = − ⎝ ⎠ =
xm n
∞
e −t
∫
βz m −1
− z= ,
n t z +1 n
βx n

⎛ 1 ⎞u ⎛ 1 ⎞
− μx
2 Γ⎜ − , μx 2 ⎟ − 2 π − Γ⎜ − , μu 2 ⎟
u e ⎝ 2 ⎠ ⎝ 2 ⎠
∫ dx = − =
0 x2 −
1
−
1

2μ 2
2μ 2
0
⎛ 1 ⎞
= − μπ − μπ Γ⎜ − , μu 2 ⎟
⎝ 2 ⎠
⎡ 1⎤
with parameters ⎢ β = μ , m = n = 2, z = ⎥ ,
⎣ 2⎦
⎛ 1 ⎞ ⎛ 1⎞
and Γ⎜ − ,0 ⎟ = Γ⎜ − ⎟ = −2 π by Formula
⎝ 2 ⎠ ⎝ 2⎠
8.338 3.
Then, from Sect. 8.359, below,
⎛ 1 ⎞ ⎡ e−x ⎛ 1 ⎞⎤
Γ⎜ − , x ⎟ = 2 ⎢ − Γ⎜ , x ⎟⎥ , so that
⎝ 2 ⎠ ⎣ x ⎝ 2 ⎠⎦
⎡ e−μu ⎞⎤
2

⎛ 1 ⎞ ⎛1
Γ⎜ − , μu 2 ⎟ 2⎢ − Γ⎜ , μu 2 ⎟⎥
⎝ 2 ⎠ = ⎣⎢u μ ⎝2 ⎠⎦⎥ 1 −μu2 ⎛1 ⎞
= e − μΓ⎜ , μu 2 ⎟
⎝ ⎠
1 1
− − u 2
2μ 2 2μ 2
and applying Formula of Sect. 8.359, below,
⎛1 ⎞
Γ⎜ , x ⎟ = π − π Φ x , then, ( )
⎝2 ⎠
⎛ 1 ⎞
Γ⎜ − , μu 2 ⎟
⎝ 2
−
1
⎠ = 1 e − μu 2 − μπ 1 − Φ u μ
u
[ ( )]
2μ 2
so that,
2
u
∫
e− μx
0 x
2
⎧1 −μu2
dx = − ⎨ e
⎩u
⎫
− μπ 1 − Φ u μ ⎬ =
⎭
[ ( )]
1 −μu2
− e
u
− μπ + μπ 1 − Φ u μ = [ ( )]
1 −μu2
− e
u
− μπ Φ u μ . [ ( )]
With the same approach,
2
ue− βx 1 −βu2
∫ 2 dx = − e
0 x u
− βπ Φ u β [ ( )]
so that,
u
⎛ − μx 2 − βx 2 ⎞ dx
∫ ⎜e −e ⎟ 2=
0⎝ ⎠x
− βu 2
− μu 2
e −e
u
+ βπ Φ u β − μπ Φ u μ [ ( )] [ ( )]

NOTE: As a check, the calculation, for

——Sect. 3.464——
 149

u → ∞, Φ (∞ ) = 1, yields,
∞
⎛ − μx 2
∫ ⎜e
0⎝
⎞
2 dx
− e− βx ⎟ 2 = β π − μ π = π
⎠x
( )
β− μ,

Formula 3.464
¾ Verified on math.com

——Sect. 3.464——
 150

FORMULA PROOF OUTLINE & NOTES

(formula references are to G & R 7e)
⎛ b ⎞
v By Formula in Sect. 3.326,
∫ exp⎜⎝ − x n ⎟
⎠
dx = v
( ) Γ(γ , βu n ) − Γ(γ , βv n ) m +1
∫u − β = γ =
m n
u x exp x dx γ
1 nβ n
bn ⎡ ⎛ 1 b ⎞ ⎛ 1 b ⎞⎤ with parameters
⎢ Γ⎜ − n , v n ⎟ − Γ ⎜ − n , u n ⎟⎥
⎡ ⎤
n ⎣ ⎝ ⎠ ⎝ ⎠⎦ 1
⎢m = 0, β = b, n = − n, γ = − n , u = u , v = v ⎥
⎛ b ⎞ ⎛ b ⎞ ⎣ ⎦
= v exp⎜ − n ⎟ − u exp⎜ − n ⎟ Second form based on substituting
⎝ v ⎠ ⎝ u ⎠
Formula 8.356.2, Γ(α , x ) with α = − into first
1
⎡ ⎛ 1 b ⎞ ⎞⎤
1
⎛ 1 b
+ b n ⎢Γ⎜1 − , n ⎟ − Γ⎜1 − , n ⎟⎥ n
⎣ ⎝ n u ⎠ ⎝ n v ⎠⎦ form.
[Re n > 0, Re b > 0] ¾ Verified on math.com

u
⎛ b ⎞ • See Bers, vol. 1, Sect. 8.5, p. 404
∫ exp⎜⎝ − x
0
n ⎟
⎠
dx = (“Transcendental Functions”), Theorem 1
and Sect. 5.3, p. 130 (“Limits at infinity”)
1
⎛ b ⎞ ⎛ 1 b ⎞ • Assume model of a continuous function,
u exp⎜ − n ⎟ − b n Γ⎜1 − , n ⎟
⎝ u ⎠ ⎝ n u ⎠ f (x);
[u > 0, n > 0, Re b > 0] ⎧ − xbn
⎪ x≠0
f ( x) = ⎨e
⎪⎩0 x=0
Then, let

1
b ⎛ b ⎞ n dt nb ⎛ b ⎞
t = n , x = ⎜ ⎟ , = − n+1 , t = lim⎜ n ⎟ = ∞,
x ⎝ t ⎠ dx x 0 ⎝ ⎠
x →0 x

b
tu = n ,
u
to give,
1 b
n
n u − 1 −1
b
∫t
−t
I =− n
e dt =
n ∞
1 1
∞ 1
bn − −1 bn ⎛ 1 b ⎞
∫t
−t
n
e dt = Γ⎜ − , ⎟
n b n ⎝ n un ⎠
un

by Formula 8.350 2, Γ(α , x ).

Final form is,

—Sect. 3.471—
 151

bn ⎛ 1 b ⎞
I= Γ⎜ − , ⎟ =
n ⎝ n un ⎠
1
⎛ b ⎞ ⎛ 1 b ⎞
u exp⎜ − n ⎟ − b n Γ⎜1 − , n ⎟
⎝ u ⎠ ⎝ n u ⎠
by Formula 8.356.2, Γ(α , x ), by setting α = − .
1
n
¾ Math.com gives different answer
u
⎛ b⎞ ⎛ b⎞ ⎛ b⎞
u
By above integral, ∫ exp⎛⎜⎜ − ⎞
∫0 exp⎜⎝ − x ⎟⎠dx = u exp⎜⎝ − u ⎟⎠ + b Ei ⎜⎝ − u ⎟⎠
b ⎟dx,
⎟
⎝ xn ⎠
0

[u > 0, Re b > 0] with parameter n = 1 , and Formula 8.359 1,

Γ(0, x ) .

¾ Verified on Math.com
⎛ b ⎞
u
⎛ b ⎞
u

∫ exp⎜⎝ − x
0
2 ⎟
⎠
dx = By above integral, ∫ exp⎜ − n ⎟dx , with parameter
0 ⎝ x ⎠
⎛ b ⎞ ⎡ ⎛ b ⎞⎤ ⎛1 ⎞
u exp⎜ − 2 ⎟ − bπ ⎢1 − Φ⎜⎜ ⎟⎥ n = 2 , and Γ⎜ , x ⎟, Sect. 8.359, below.
⎝ u ⎠ ⎟ ⎝2 ⎠
⎢⎣ ⎝ u ⎠⎥⎦
¾ Verified on Math.com
[u > 0, Re b > 0]
u
⎛ β ⎞ dx 1 ⎛ β ⎞ • In integral, set
∫0 exp⎜⎝ − x n ⎟⎠ x n+1 = nβ exp⎜⎝ − u n ⎟⎠ β
1
⎛ β ⎞ n ds nβ ⎛β ⎞
s = n , x = ⎜ ⎟ , = − n+1 , s 0 = lim⎜ n ⎟ = ∞,
[u > 0, Re β > 0 Re n > 0] x ⎝ s ⎠ dx x x → 0
⎝x ⎠
β
su = ,
un

to give,
β
un
x n +1 − s
∫
1
I =− e ds =
nβ x n +1
∞

⎛ β ⎞
∞ Γ⎜1, n ⎟
e − s ds = ⎝ ⎠,
1 u
∫
nβ β nβ
un

by Formula 8.356.2, Γ(α , x ), and apply Formula,

n −1
xm
Γ(n, x ) = (n − 1)! e − x
∑ [n = 1,2,...] .
m=0 m!

• Or, in the integral below,

⎛ b ⎞ ⎛ m −1 b ⎞
u exp⎜ − n ⎟ Γ⎜ , n⎟
⎝ x ⎠ dx = ⎝ n u ⎠
∫0 x m m −1 ,
n
nb

—Sect. 3.471—
 152

set b = β , m = n + 1, and apply Formula,

n −1
xm
Γ(n, x ) = (n − 1)! e − x ∑ [n = 1,2,...] .
m = 0 m!

Example: In Formula, 3.471.1, if n = 1 ,

⎛ β ⎞ dx 1 ⎛ β ⎞
u

∫0 exp⎜⎝ − x ⎟⎠ x 2 = β Γ⎜⎝1, u ⎟⎠ =
.
1 ⎛ β⎞
exp⎜ − ⎟
β ⎝ u⎠
¾ Verified on Math.com
• Change of variable,
∫ exp[− β (ax + b ) ]dx
v
−n
ds
u s = ax + b, = a, su = au + b,
1 dx
β n ⎡ ⎛⎜ 1 β ⎞ ⎛ 1
⎟ − Γ⎜ − ,
β ⎞ s v = av + b, giving,
= ⎢Γ⎜ − ,
na ⎢⎣ ⎝ n (av + b ) ⎟⎠ n ⎜ n (au + b )n
⎝ ⎠
s
1 v −n
β
I = ∫ e s ds
⎧ ⎫ a su
⎪ ⎪
⎪ ⎪
⎪ ⎪ o Use Sect. 3.326 integral, above,
⎡
⎪exp − β ⎤ ⎪ Γ(γ , βun ) − Γ(γ , βvn ) m +1
⎪ ⎢ (av + b )n ⎥ (av + b ) − ⎪
v m
x exp (− βx n
)dx = ,γ =
⎪ ⎣ ⎦ ⎪ ∫u nβ γ
n

1 ⎪⎪ ⎡ β ⎤ ⎪⎪
= ⎨exp ⎢− ⎥ ( au + b ) ⎬ ⎡ 1⎤
a ⎪ ⎣ (au + b )n ⎦ ⎪ with parameters ⎢m = 0, n = − n, β = β , γ = − ⎥
⎪ ⎪ ⎣ n⎦
⎡ ⎛ β ⎞ ⎤
⎪ ⎟− ⎪ • Second form based on Formula 8.356.2,
1
⎢Γ⎜⎜1 − , n ⎟ ⎥⎪
⎪ n (au + b ) ⎠ ⎥
⎪+ β n ⎢ ⎝
1
Γ(α , x ) , with α = − .
1
⎪
⎢ ⎥
⎪
⎢ ⎛ 1 β ⎞ ⎥⎪ n
⎪ Γ ⎜ 1 − , ⎟ ⎪ ¾ Verified on Math.com
⎪⎩ ⎢ ⎜⎝ n (av + b )n ⎟⎠ ⎥ ⎪
⎣ ⎦⎭

[Re β > 0, Re n > 0, Re a > 0]

—Sect. 3.471—
 153

[ ]dx = In above integral,

u

∫ exp − β (ax + b)
−n

∫ exp[− β (ax + b) ]dx =

v
−n
0
u
1

β ⎡ ⎛ 1 ⎞⎤
1
β ⎞ 1 β
⎟ − Γ⎛⎜ − , n βn ⎡ ⎛ 1
n
β ⎞ ⎛ 1 β ⎞⎤
⎢Γ⎜⎜ − , n ⎟ ⎟⎥ ⎢Γ ⎜ − , ⎟ − Γ⎜⎜ − ,
na ⎢⎣ ⎜⎝ n (av + b )n ⎟⎠
⎟
n ⎟⎥
na ⎣⎢ ⎝ n (au + b ) ⎠ ⎝ n b ⎠⎦⎥ ⎝ n (au + b ) ⎠⎦⎥
set u = 0, v = u
⎧ ⎫
⎪ ⎪ • Second form based on Formula 8.356.2,
⎪ ⎪
Γ(α , x ) , with α = − .
1
⎪ ⎡ β ⎤ ⎪
⎥ (au + b )
⎪exp ⎢− ⎪ n
⎪ ⎣ (au + b )n ⎦ ⎪ ¾ Verified on Math.com
⎪ ⎪
1⎪ ⎡ β⎤ ⎪
= ⎨− b exp ⎢− n ⎥ ⎬
a⎪ ⎣ b ⎦ ⎪
⎪ ⎡ ⎛ 1 β ⎞ ⎤⎪
⎪ 1 ⎢
Γ⎜1 − , n ⎟ ⎥⎪
⎪ n⎢
⎝ n b ⎠ ⎥⎪
+ β
⎪
⎪
⎢ ⎛ 1 β ⎞⎥ ⎪⎪
⎢− Γ⎜⎜1 − , ⎟⎥
⎟ ⎪
⎪ ⎢⎣ ⎝ n (au + b )
n
⎩ ⎠⎥⎦ ⎭

[u > 0, Re β > 0, Re n > 0, Re a > 0, Re b > 0]

∫ exp[− β (ax + b ) ]dx =

u
−1
In above integral (second form),
∫ exp[− β (ax + b ) ]dx
u
−n ,
0
0

⎧ ⎛ β ⎞ ⎛ β ⎞ ⎫ set n = 1, and Formula 8.359.1, Γ(0, x ).

⎪(au + b ) exp⎜ − ⎟ − b exp⎜ − ⎟ + ⎪
1⎪ ⎝ au + b ⎠ ⎝ b⎠ ⎪
⎨ ⎬ ¾ Verified on Math.com
a⎪ ⎡ ⎛ β ⎞ ⎛ β ⎞⎤ ⎪
β ⎢Ei⎜ − ⎟ − Ei⎜ − ⎟⎥
⎪⎩ ⎣ ⎝ au + b ⎠ ⎝ b ⎠⎦ ⎪⎭

[u > 0, Re β > 0, Re a > 0, Re b > 0]

[ ]
u
In above integral (second form),
∫ exp − β (ax + b ) dx =
−2

∫ exp[− β (ax + b ) ]dx

u
−n ,
0
0

⎧ ⎡ ⎤ ⎫ ⎛1 ⎞
⎪(au + b ) exp ⎢− β ⎪ set n = 2 , and Formula, Γ⎜ , x ⎟, Sect 8.359,
2 ⎥ ⎝2 ⎠
⎪ ⎣ (au + b ) ⎦ ⎪
⎪ ⎪ below.
1⎪ ⎛ β ⎞ ⎪ ¾ Not verifiable on Math.com
⎨− b exp⎜ − 2 ⎟ + ⎬
a⎪ ⎝ b ⎠ ⎪
⎪ ⎡ ⎛ ⎞ ⎛ ⎞⎤ ⎪
⎪ βπ ⎢Φ⎜ β ⎟ − Φ⎜ β ⎟⎥ ⎪
⎪
⎩ ⎢⎣ ⎜⎝ au + b ⎟⎠ ⎜ b
⎝
⎟⎥ ⎪
⎠⎦ ⎭

[u > 0, Re β > 0, Re a > 0, Re b > 0]

—Sect. 3.471—
 154

u
⎡ β (ax + b ) ⎤ In above integral (second form),
∫0 exp⎢− ax + b ⎥ dx ∫ exp[− β (ax + b ) ]dx
u
−n ,
⎣ ⎦ 0

⎡u > 0, Re β > 0, Re a > 0, Re b > 0,⎤

, and apply Formula 8.352.3, Γ(− n, x ) .
1
⎢Re(b + au ) > 0 ⎥ set n =
⎣ ⎦ 2
¾ Verified on Math.com

∫ exp[− β (b − ax ) ]dx = [ ]
v v
−n Similar to above integral, ∫ exp − β (ax + b )−n dx
u
u
1 ¾ Verified on Math.com
βn ⎡ ⎛ 1 β ⎞ ⎛ 1 β ⎞⎤
⎢Γ ⎜ − , ⎟ − Γ⎜ − , ⎟⎥
na ⎢⎣ ⎜⎝ n (b − au )n ⎟
⎠
⎜ n (b − av )n
⎝
⎟
⎠⎥⎦
⎧ ⎫
⎪ ⎪
⎪ ⎪
⎪ ⎪
⎪b exp ⎡− β ⎤
(b − au ) − ⎪⎪
⎪ ⎢ n ⎥
⎪ ⎣ (b − au ) ⎦ ⎪
⎪
1⎪ ⎡ β ⎤ ⎪⎪
= ⎨exp ⎢− ⎥ (b − av ) ⎬
a ⎪ ⎣ (b − av )n ⎦ ⎪
⎪ ⎪
⎪ ⎡ ⎛ 1 β ⎞ ⎤⎪
⎢Γ⎜⎜1 − , ⎟−
n ⎟ ⎥⎪
⎪ ( − )
⎪+ β n ⎢ ⎝
1 n b av ⎠ ⎥⎪
⎢ ⎥
⎪
⎢ ⎛ 1 β ⎞ ⎥⎪
⎪ Γ ⎜ 1 − , ⎟ ⎪
⎪⎩ ⎢ ⎜⎝ n (b − au )n ⎟⎠ ⎥ ⎪
⎣ ⎦⎭

[Re β > 0, Re n > 0, Re a > 0]

∫ exp[− β (b − ax ) ]dx In above integral, exp[− β (b − ax )− n ]dx , set
u v
−n
∫ u
0
1
u = 0, v = u.
β ⎡ ⎛ 1 β ⎞
n ⎛ 1 β ⎞⎤
= ⎢Γ⎜ − , n ⎟ − Γ⎜⎜ − , ⎟⎥
⎟ ¾ Verified on Math.com
na ⎣⎢ ⎝ n b ⎠ ⎝ n (b − au )
n
⎠⎦⎥
⎧ ⎫
⎪ ⎪
⎪ ⎪
⎪ ⎪
⎪b exp ⎡− β ⎤ − ⎪
⎪ ⎢⎣ b n ⎥⎦ ⎪
⎪ ⎪
1⎪ ⎡ β ⎤
= ⎨exp ⎢− n ⎥
(b − au ) ⎪⎬
a ⎪ ⎣ (b − au ) ⎦ ⎪
⎪ ⎪
⎪ ⎡ ⎛ 1 β ⎞ ⎤⎪
1⎢ ⎜
Γ⎜1 − , ⎟ −⎥
⎪
n⎢ ⎝
n (b − au )n ⎟⎠ ⎥ ⎪
⎪+ β ⎪
⎪ ⎢ 1 β ⎞ ⎥⎪
⎛
⎢Γ⎜1 − , n ⎟ ⎥⎪
⎪
⎩ ⎢
⎣ ⎝ n b ⎠ ⎦⎥ ⎭
—Sect. 3.471—
 155

[u > 0, Re β > 0, Re n > 0, Re a > 0, Re b > 0]

Set n = 1 in above integral (second form),
[ ]
u

∫ exp − β (b − ax ) dx =
−1

∫ exp[− β (b − ax ) ]dx , and apply Formula

u
−n
0

⎛ β⎞ β ⎞⎫
0
⎧ ⎛
⎪b exp⎜ − b ⎟ − (b − au ) exp⎜ − b − au ⎟⎪ 8.359.1, Γ(0, x )
1⎪ ⎝ ⎠ ⎝ ⎠⎪ ¾ Verified on Math.com
⎨ ⎬
a⎪ ⎡ ⎛ β⎞ ⎛ β ⎞⎤ ⎪
+ β ⎢Ei⎜ − ⎟ − Ei⎜ − ⎟⎥
⎪⎩ ⎣ ⎝ b⎠ ⎝ b − au ⎠⎦ ⎪⎭
[u > 0, Re β > 0, Re a > 0, Re b > 0]
Set n = 2 in above integral (second form),
[ ]dx =
u

∫ exp − β (b − ax )
−2
⎛1 ⎞
∫0 exp[− β (b − ax ) ]dx ,and Formula, Γ⎜⎝ 2 , x ⎟⎠, Sect.
u
−n
0

⎧ ⎛ β ⎞ ⎛ β ⎞⎫
⎪b exp⎜ − 2 ⎟ − (b − au ) exp⎜⎜ − ⎟⎪ 8.359, below.
⎟
⎝ b ⎠ ⎝ (b − au )
2
⎪ ⎠⎪ ¾ Verified on Math.com
⎪ ⎪
1⎪ ⎡ ⎛ β⎞ ⎤ ⎪
⎨ ⎢Φ⎜⎜ ⎟ ⎥ ⎬
a⎪ ⎢ ⎝ b ⎟⎠ ⎥ ⎪
⎪ + βπ ⎢ ⎥ ⎪
⎢ ⎛ β ⎞
⎪ − Φ⎜ ⎟⎥ ⎪
⎪ ⎢ ⎜ ⎟⎥ ⎪
⎩ ⎣ ⎝ b − au ⎠⎦ ⎭

[u > 0, Re β > 0, Re a > 0, Re b > 0]

u
⎡ β ⎤ Set n =
1
∫0 exp⎢⎣− b − ax ⎥⎦dx 2
in above integral (second from),

exp[− β (b − ax ) ]dx , and apply Formula 8.352.3,

u

∫
−n
⎧ ⎫
⎪ ⎪ 0

⎪ ⎪ Γ(− n, x ) .
⎪ ⎡ ⎤⎡ ⎤⎪ ¾ Verified on Math.com
⎪exp⎢− β ⎥ ⎢β b − au ⎥⎪
⎪ ⎣ b − au ⎦ ⎣− (b − au) ⎦⎪
⎪ ⎪
1⎪ ⎛ β ⎞
= ⎨− exp⎜⎜ − (
⎟⎟ β b − b
⎪
⎬ )
a⎪ ⎝ b⎠ ⎪
⎪ ⎡ ⎛ ⎪
⎪ β ⎞⎤ ⎪
⎢ Ei ⎜ − ⎟⎥
⎪ 2 ⎢ ⎜⎝ b − au ⎟⎠⎥ ⎪
⎪+ β ⎢ ⎪
⎪ ⎛ β ⎞ ⎥ ⎪
⎢− Ei⎜⎜ − ⎟⎟ ⎥
⎪ ⎢⎣ ⎪
⎩ ⎝ b ⎠ ⎥⎦ ⎭

⎡u > 0, Re β > 0, Re a > 0, Re b > 0,⎤

⎢Re(b − au ) > 0 ⎥
⎣ ⎦
—Sect. 3.471—
 156

⎡ b ⎤
u
In above integral,
∫0 exp⎢⎣− a + x ⎥⎦ dx = ∫ exp[− β (ax + b)
u
−1
]dx ,
0

(a + u ) exp⎛⎜ − b ⎞⎟ set β = b, a = 1, b = a .
⎝ a+u⎠
⎛ b⎞ ⎡ ⎛ b ⎞ ⎛ b ⎞⎤ ¾ Verified on Math.com
− a exp⎜ − ⎟ + b ⎢Ei⎜ − ⎟ − Ei⎜ − ⎟⎥
⎝ a⎠ ⎣ ⎝ a+u⎠ ⎝ a ⎠⎦

[u > 0, Re a > 0, Re b > 0]

[
∫ exp − β (x ) dx ] Similar to above integral,
u
n+ p −1

⎛ b ⎞ ⎛ b ⎞ n ⎛ 1 b ⎞
u 1
0
∫0 ⎜⎝ x n ⎟⎠
exp − dx = u exp⎜ − n ⎟
⎝ u ⎠
− b Γ⎜ 1 − , n ⎟
⎝ n u ⎠
β ⎞ ⎡ ⎛ β
1
⎛ 1 ⎞
= u exp⎜ − n + p ⎟ − β n + p ⎢Γ⎜⎜1 − , n+ p ⎟⎟ NOTE: If p = 0, result consistent with above
⎝ u ⎠ ⎣ ⎝ n+ p u ⎠ integral,
1
⎛ b ⎞ ⎛ b ⎞ n ⎛ 1 b ⎞
u
[u > 0, Re(n + p ) > 0, Re β > 0] ∫0 ⎜⎝ x n ⎟⎠
exp − dx = u exp⎜ − n ⎟
− b Γ⎜ 1 − , n ⎟
⎝ u ⎠ ⎝ n u ⎠
¾ Not verifiable on math.com
⎛ b ⎞ ⎛ m −1 b ⎞ Let
u exp⎜ − n ⎟ Γ⎜ , n⎟ 1
⎝ x ⎠ dx = ⎝ n u ⎠ ⎛ b ⎞ n dt ⎛ b ⎞
∫0 x m m −1
b nb
t = n , x = ⎜ ⎟ , = − n+1 , t 0 = lim⎜ n ⎟ = ∞,
nb n x ⎝ t ⎠ dx x ⎝ ⎠
x →0 x

b
tu = n ,
[u > 0, Re b > 0, Re n > 0] u
to give,
b
un m −1 ∞ m −1
1 −1 1 −1
I =− m −1∫t
∞
n
e −t dt = m −1 ∫t
b
n
e −t dt
n n
nb nb un

1 ⎛ m −1 b ⎞
= m −1
Γ⎜ , n⎟
n
⎝ n u ⎠
nb
by Formula 8.350 2, Γ(α , x ).
NOTE: If m = 0, result consistent with above
integral,
1
⎛ b ⎞ ⎛ b ⎞ ⎛ 1 b ⎞
u

∫0 exp⎜⎝ − x n ⎟⎠dx = u exp⎜⎝ − u n ⎟⎠ − b n Γ⎜⎝1 − n , u n ⎟⎠

¾ Verified on Math.com
β ⎛ m −1 β ⎞ ⎛ m −1 β ⎞ By Formula 2.33.17 with n → −n
v
− Γ⎜ , n ⎟ − Γ⎜ , ⎟
⎝ n v ⎠ ⎝ n un ⎠
n
e x
∫u x m dx = m −1

nβ n

[Re β > 0, Re n > 0]

—Sect. 3.471—
 157

β ⎛ m −1⎞ By above integral,

− Γ⎜ ⎟
∞
e x n
⎝ n ⎠ m −1 β ⎛ m −1 β ⎞ ⎛ m −1 β ⎞
Γ⎜ , ⎟ − Γ⎜
∫0 x m dx = m −1
, >0 v
e
− n
x
⎝
, ⎟
n vn ⎠ ⎝ n un ⎠
nβ n
n
∫u x m dx = m −1

⎡ m −1 ⎤ nβ n
⎢Re β > 0, Re n > 0, Re n > 0⎥
⎣ ⎦
β ⎛ m −1⎞ ⎛ m −1 β ⎞ ⎛ m −1 β ⎞ By above integral,
∞
−
n Γ⎜ ⎟ − Γ⎜ , n⎟ γ⎜ , n⎟
⎝ n ⎠ ⎝ n u ⎠= ⎝ n u ⎠,
⎛ m −1 β ⎞ ⎛ m −1 β ⎞
e x
∫u x m dx = m −1 m −1 β
− n Γ ⎜ , ⎟ − Γ⎜ , ⎟
nβ n nβ v
⎝ n vn ⎠ ⎝ n un ⎠
n
e x
m −1
>0 ∫u x m dx = m −1
n nβ n
[Re β > 0, Re n > 0]
m +1
⎛ m +1 b ⎞ Similar to above integral,
b n
Γ⎜ − , ⎟ ⎛ b ⎞ ⎛ m −1 b ⎞
⎛ b ⎞
u
⎝ n un ⎠ u exp⎜ − n ⎟ Γ⎜ , n⎟
∫x exp⎜ − n ⎟dx =
m
⎝ x ⎠ dx = ⎝ n u ⎠
0 ⎝ x ⎠ n
∫0 x m m −1
nb n
[u > 0, Re b > 0, Re n > 0]
NOTE: If m = 0, result consistent with above
integral,
⎛ b ⎞
1
⎛ b ⎞ n ⎛ 1 b ⎞
u

∫0 ⎝ x n ⎠
exp⎜ − ⎟ dx =u exp⎜ −
⎝ u ⎠
n ⎟ − b Γ⎜ 1 − , n ⎟
⎝ n u ⎠
¾ Verified on Math.com
v
(a − x )μ −1 exp⎛ − β ⎞dx = A more challenging integral
∫ x μ +1
⎜
⎝
⎟
x⎠ • Change of variable:
u
β ds β β β
−
β Set s = ; = − 2 ; su = ; sv = to give
a μ −1e ⎡ ⎛ β β⎞ a
⎛ β β ⎞⎤ x dx x u v
⎢ Γ⎜ μ , − ⎟ − Γ⎜ μ , − ⎟ ⎥ μ −1
β μ
⎣ ⎝ v a⎠ ⎝ u a ⎠⎦ ⎛ β⎞
2 sv ⎜ a − ⎟
[Re β > 0, Re a > 0] x ⎝
I =− ∫
s⎠
e − s ds =
β su x μ +1
μ −1
⎛ β⎞
su ⎜ a − ⎟ s
1 ⎝ s⎠ 1 u
∫ e ds = μ ∫ (as − β ) e − s ds
−s μ −1
μ −1
β sv ⎛ β ⎞ β sv
⎜ ⎟
⎝s⎠
o 2nd step; now use the integral derived in
Sect. 3.462,
v

∫ (ax ± b ) e − ρx dx =
m

u
ρb
±
a e m
⎡ ⎛ a
ρb ⎞ ⎛ ρb ⎞⎤
⎢ Γ⎜ m + 1, ρu ± ⎟ − Γ⎜ m + 1, ρv ± ⎟
ρ m +1
⎣ ⎝ a ⎠ ⎝ a ⎠⎥⎦
with parameters,
[a = a, b = β , m = μ − 1, ρ = 1, u = sv , v = su ] to get,
—Sect. 3.471—
 158

⎡ ⎛ β⎞ ⎤
⎢Γ⎜ μ , sv − a ⎟ ⎥
β
−
μ −1
⎢ ⎝ ⎠ ⎥
a
a e
I= which leads to
βμ ⎢ ⎛ β ⎞⎥
⎢− Γ⎜ μ , su − ⎟⎥
⎣ ⎝ a ⎠⎦
solution.
o NOTE: Formula 3.471.3 results if
u = 0, v = u = a noting that
β β
sv = = +∞; Γ(a,0) = Γ(a ), Γ(a, ∞ ) = 0.
, su = lim
a u →0 u

NOTE: How else can the second step be done?

—Sect. 3.471—
 159

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
Formula 3.473 1 Constraint added since, by Formula 3.362.2,
∞
(2m − 1)!! π
∫ exp(− x )x
( m +1 2 )n −1
integral is equal to,
n
dx = m ⎛ 1⎞
0 2 n Γ⎜ m + ⎟
I= ⎝
[n > 0, m = 0,1,2,...] 2⎠
. We assume author then
n
⎛ 1⎞
employed integer-formula 8.339.2, Γ⎜ m + ⎟
⎝ 2⎠
which requires m ≥ 0.
∞
(2m − 3)!! π
∫ exp(− βx )x
( m −1 2 )n −1
Use Formula 3.362.2 with parameters,
n
dx = ⎡ n ⎤
⎢m = mn − − 1, n = n, β = β ,⎥
1
m−
m −1
nβ
0 2 ⎢ 2 ⎥ .
2 ⎢ 1 ⎥
γ = −
[Re β > 0, n > 0, m = 1,2,...] ⎢ m ⎥
⎣ 2 ⎦

⎛ 1⎞
Γ⎜ m − ⎟
Then, I = ⎝
2⎠
1
, where
m−
2 m −1 nβ 2

⎛ 1⎞ π (2m − 3)!!
Γ⎜ m − ⎟ = , m ≥ 1.
⎝ 2⎠ 2 m −1
by formula in Sect. 8.339, below.
¾ Verified on math.com
∞
(− 1)m 2 m
∫ exp(− βx )x
(1 2−m )n −1
Similar to above integral,
n
dx = π ∞ ⎛ 1⎞
(2m − 3)!!
∫ exp(− βx )x
⎜⎜ m − ⎟⎟ n − 1
1
−m
n ⎝ 2⎠
dx = π
(2m − 1)!!
1

nβ
m−
0 2 0
2 m −1 nβ 2

[Re β > 0, n > 0, m = 1,2,...]

⎛1 ⎞
with numerator, Γ⎜ − m ⎟ , Formula 8.339.3
⎝2 ⎠
¾ Verified on math.com

—Sect. 3.473—
160
161
 162

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
= m Ei(ln u ) ± Ei(ln v ) = m li(u ) ± li(v )
v dx v
= Ei(ln v ) − Ei(ln u ) by change of
dx
∫ • ∫
u ± ln x ln x
u

variable,
ds 1
s = ln x, = , su = ln u, sv = ln v,
dx x
e s = e ln x = x.
v sv s
sv
= ∫ ds = Ei(s ) = Ei(ln v ) − Ei(ln u )
dx e
I =∫
u
ln x su s su
= li(v ) − li(u )
by integral 2.325.7,
e ax
n
Ei ax n ( )
∫ x dx = n and Formula 8.240.1.
• Likewise,
v
= Ei(ln u ) − Ei(ln v )
dx dx
v
= −∫
∫ − ln x u ln x
= li(u ) − li(v )
u

NOTE: if u = 0, v = u , su = ln(0) = −∞, sv = ln(v),

then,
u

∫ = Ei(ln u ) − Ei(ln 0 ) = Ei(ln u ) = li(u ),

dx
ln x
0
Formula 4.211 2 by Formula 8.240 1.

¾ Verified on math.com
∞ Follows from above integral,
dx
∫ ln x = +∞
e
v

∫ ln x = Ei(ln v ) − Ei(ln u ) = li(v ) − li(u ),

u
dx

with u = e, v = ∞.
Also, use change of variable,
ds 1
s = ln x, = , se = ln e = 1, s∞ = ∞,
dx x
e = e = x.
s ln x

∞ ∞ s
I = ∫ ds = = ∫ ds =Ei(s ) =
x e ∞

1
s 1
s 1

Ei(∞ ) − Ei(1) = +∞
since Ei(1) ≈ 1.895117.... by 8.218.2
¾ Not verifiable on math.com. In addition,
published result, Formula 4.211.1,

—Sect. 4.211—
 163

∞
dx
∫e ln
1
= −∞ , was not verifiable on math.com, but

x
its indefinite form was returned,
⎛ 1⎞
∫ 1 = −Ei⎜ − log ⎟ = − Ei(log x ) ,
dx
ln ⎝ x⎠
x
which produces,
∞
∫ 1 = −Ei(log x ) e = −∞.
dx
ln
x
Likewise, calculation for indefinite form,
∞

∫e ln x = +∞ , returns ∫ ln x = li(x ) which produces,

dx dx

from Formula 8.240 1,

∞ ∞
dx
∫ ln x = li ( x ) = Ei( ln x ) =
e e
Ei(ln ∞) − Ei(ln e) = Ei( ∞) − Ei(1) = +∞.
NOTE: similar calculations show, based on
properties,
v s
u
∫u s ds = Ei(s ) v = Ei(u ) − Ei(v )
e

o Ei(ln ∞ ) = Ei(+ ∞ ) = ∞,
Ei(ln 0) = Ei(− ∞ ) = 0,
Ei(ln1) = Ei(0) not defined.
that,
∞ ∞ ∞
dx dx dx
o ∫ = ∫ = ∫ = −∞
0 ln
1 1 ln
1 e ln
1
x x x
∞ ∞ ∞
dx dx dx
o ∫0 ln x = ∫1 ln x = ∫e ln x = +∞
v
• Indefinite integral solution by change of
∫ ln xdx = u(1 − ln u ) − v(1 − ln v )
u
variable, s = ln x :
∫ x ln xdx = x(ln x − 1) = − x(1 − ln x ) =
v v
u u
u (1 − ln u ) − v(1 − ln v ).

• Or—By change of variable on definite integral:

ds 1
s = ln x, = , su = ln u , sv = ln v,
dx x
e s = e ln x = x

—Sect. 4.211—
 164

sv sv
sv
I = ∫ xsds = ∫ se s ds = −Γ(2,− s ) =
su su
su
Γ(2,− ln u ) − Γ(2,− ln v )
= e ln u (1 − ln u ) − e ln v (1 − ln v )
= u (1 − ln u ) − v(1 − ln v )
by integral previously submitted,
( )1−γ Γ γ ,− ax
n
m +1 ( )
∫ = − ,γ =
m ax n
x e 1 γ
na n
with parameters [m = 1, a = n = 1, γ = 2] ,
n −1
xm
and Γ(n, x ) = (n − 1)! e − x ∑
m =0 m!

⎧− ln u
with n = 2, x = ⎨
⎩− ln v
NOTE: Test—if
u = 1, ln(1) = 0, v = e, ln(e) = 1,
e

∫ ln xdx = 1(1 − 0) − e(1 − 1) = 1

1
∞
Corollary: ∫ ln xdx = 0 − ∞(1 − ∞) = +∞.
0

¾ verified math.com
v Change of variable,
∫ x ln x = ln(ln v ) − ln(ln u )
dx
∫ = ln(s ) = ln(ln x )
ds 1 ds v
s = ln x, = ;I =
dx x s u
.
u
¾ Verified on math.com

( ) ( )
v n
x dx From Formula 2.724.2,
∫ = li v n+1 − li u n+1 x n dx n +1 v
( ) ( ) ( )
u
ln x
∫ ln x = li x
u
= li v n+1 − li u n+1

¾ Verified on math.com

( ) ( ) ( )
u v
x n dx x n dx
∫
0
ln x
= li u n +1 In ∫
u
ln x
= li v n+1 − li u n+1 ,

set u = 0, ln(0) = −∞, v = u, ln(v) = ln u.

Then,
I = Ei[(n + 1)s ] = Ei[(n + 1) ln u ] − 0
ln u
−∞

= li⎛⎜ u n +1 ⎞⎟.
⎝ ⎠

¾ Not verifiable on math.com

—Sect. 4.211—
 165

( ) ( )
v Indefinite integral solution:
∫u ln x = Ei(2 ln v ) −Ei(2 ln u ) = li v − li u
xdx 2 2
ds 1
Set s = ln x, = , su = ln u , sv = ln v, e s = x.
[2 ln u > 0,2 ln v > 0] dx x
e2s
Then, I = ∫ ds = Ei(2s ) = Ei(2lnx) , or
s
v

∫ = Ei(2 ln x ) =
xdx u
ln x v
u
Ei(2 ln v ) − Ei(2 ln u ) = li⎛⎜ e 2 ln v ⎞⎟ − li⎛⎜ e 2 ln u ⎞⎟ =
⎝ ⎠ ⎝ ⎠
⎛ 2 ⎞ ⎛ 2
li⎜ v ⎟ − li⎜ u ⎟⎞
⎝ ⎠ ⎝ ⎠

by Formula 8.240.1, Ei( x ) = li e x , and relation ( )

e =a .
x ln a x

NOTE: cf. Formula 2.724.2.

¾ Verified on math.com
v
v2 ⎛ 1 ⎞ u2 ⎛ 1⎞ ¾ Verified on math.com
∫u x ln xdx = 2 ⎜⎝ ln v − 2 ⎟⎠ − 2 ⎜⎝ ln u − 2 ⎟⎠
e In above integral
∫ ln x dx = 1 v
1
∫ ln xdx = u (1 − ln u ) − v(1 − ln v ) ,
u

set u = 1, ln(1) = 0, v = e, ln(e) = 1.

v v

∫ ln(ln x )dx = v ln(ln v ) − u ln(ln u ) − [li(v ) − li(u )] •

u
Change of variable on ∫ ln (ln x )dx :
u

ds 1
s = ln x, = , su = ln u, sv = ln v
dx x
e =x
s

sv

I = ∫ ln s e s ds.
su
sv

• ∫ ln s e ds :
s
Integration by parts on
su

ds
u = ln s, du = , dv = e s ds, v = e s ,
s
sv
es
uv − ∫ vdu = ln (s )e − ln (s )e − ∫ s s
ds.
su
s
• Re-express:
sv
I = ln (ln v )e ln v − ln (ln u )e ln u − Ei(s )
su
= v ln (ln v ) − u ln (ln u ) − [Ei(ln v ) − Ei(ln u )]
= v ln (ln v ) − u ln (ln u ) − [li(v ) − li(u )]
¾ Verified math.com

—Sect. 4.211—
 166

v
(ln x ) p dx = (ln v ) p +1 − (ln u ) p +1
∫
u
x p +1
¾ Verified on math.com
e
ln xdx 1 • In above integral,
∫1 x = 2 v
(ln x ) p dx = (ln v ) p+1 − (ln u ) p+1 ,
∫u x p +1
set p = 1, u = 1, ln(1) = 0, v = e, ln(e) = 1.
• Or, use change of variable,
ds 1
s = ln x, = , s1 = 0, se = ln(e) = 1,
dx x
e s = e ln x = x, to give :
1 1
xs s2 1 1
I =∫ ds = ∫ sds = = .
0
x 0
2 0 2
¾ Verified on math.com
e
(ln x )p dx = 1
In above integral, ∫
v
(ln x ) p dx = (ln v ) p +1 − (ln u ) p +1 ,
∫
1
x p +1 u
x p +1
set u = 1, ln(1) = 0, v = e, ln(e) = 1.

—Sect. 4.211—
 167

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
1 Change of variable,
ln xdx
∫ a + ln x = 1 − ae
−a
Ei(a ) [a > 0] ds 1
0 s = a + ln x, = , s0 = −∞, s1 = a,
dx x
es
e s = e a e ln x = e a x ⇒ x = a ,
e
ln x = s − a
1 (s − a )e s
a
I= a ∫ ds =
e −∞ s
1 ⎡ s es ⎤
a a

e a ⎣ −∫∞ ∫− ∞ s ds ⎥⎦ =
⎢ e ds − a

e
[
1 a
a
]
e − aEi(a ) = 1 − ae − a Ei(a )
by Formula 8.211.2, Ei( x)
¾ Verified on math.com
1 Proof similar to above integral,
ln xdx
∫ a − ln x = −1 − ae Ei(−a)
a
[a > 0] 1
ln x
∫0 a + ln xdx = 1 − ae Ei(a)
−a
0

¾ Verified on math.com
v
dx Change of variable:
∫ (a + ln x )
u
2
s = a + ln x,
ds 1
= , su = ln u + a,
dx x
+ e −a [Ei(a + ln v ) − Ei(a + ln u )]
u v
= − es
a + ln u a + ln v sv = ln v + a, e = es ln x + a
= xe ⇒ x = a
a

[a + ln u > 0, a + ln v > 0] e
Then,
sv sv
x 1 es
I = ∫ 2 ds = a ∫ s 2 ds
su
s e su

Next,
sv
I = a Γ(− 1,− s ) =
1
e su
1
[Γ(− 1,− sv ) − Γ(− 1,− su )].
ea
by indefinite integral 2.325 6,
e ax
n
(− 1) z +1
(
a z Γ − z ,−ax n ) m −1
∫ x m dx = n
z=
n
with parameters [m = 2, a = n = 1, z = 1] .
Calculations for each term show,

—Sect. 4.212—
 168

1⎡ e −(− sv ) ⎤
Γ(− 1,− sv ) = (− 1) ⎢Γ(0,− sv ) −
⎣ (− sv )1 ⎥⎦
e sv
= Ei(sv ) −
sv
by Formula 8.352.3, Γ(− n, x ), with n = 1.
Likewise,
1⎡ e −(−su ) ⎤
Γ(− 1,− su ) = (− 1) ⎢Γ(0,− su ) −
⎣ (− su )1 ⎥⎦
e su
= Ei(su ) − .
su
Hence, combining and substituting su , sv ,
⎧ e sv ⎛ e su ⎞⎫
I = e ⎨Ei(sv ) −
−a
− Ei(su ) −
⎜ ⎟⎟⎬ =
⎩ sv ⎜⎝ su ⎠⎭

⎡ e a + ln u e a + ln v ⎤
e− a ⎢ − + Ei(a + ln v ) − Ei(a + ln u )⎥
⎢⎣ a + ln u a + ln v ⎥⎦

Last, multiply in e − a and simplify.

NOTE: If
u = 0, ln(u ) = −∞, and v = 1, ln(v) = 0,
then Formula 4.212.3 results
v
dx Similar to above integral,
∫ (a − ln x )
u
2
v
dx
∫ (a + ln x ) 2
=
u

+ e a [Ei(ln v − a ) − Ei(ln u − a )]
u v ⎡ e a + ln u e a + ln v ⎤
= − e −a ⎢ − + Ei(a + ln v ) − Ei(a + ln u )⎥
ln u − a ln v − a ⎣ a + ln u a + ln v ⎦
[ln u − a > 0, ln v − a > 0] NOTE: if u = 0, ln(0) = −∞, v = 1, ln(1) = 0,
Formula 4.212.4 results.
NOTE: not verifiable on math.com, but the
indefinite integral was returned,

∫ (a − ln x )2 = e Ei(ln x − a ) − ln x − a .
dx a x

See Sect. 2.7, above, for its derivation.

v
dx Use indefinite integral solution, math.com:
∫ (a + ln x )
u
n dx
∫ (a + ln x ) n
n ⎡ v⎤
= (− 1) e − a ⎢Γ(− n + 1,−a − ln x ) ⎥ =
⎣ u ⎦
= (− 1) e −a [Γ(− n + 1,− a − ln v ) − Γ(− n + 1,−a − ln u )] (− 1)n e − a [Γ(− n + 1,−a − ln v ) − Γ(− n + 1,−a − ln u )]
n

• Or use change of variable,

ds 1
s = a + ln x, = ,
dx x
su = ln u + a, sv = ln v + a, e s = e ln x + a =
es
xe a ⇒ x =
ea
to give,
—Sect. 4.212—
 169

sv sv
x 1 es
I = ∫ n ds = a ∫ s n ds
su
s e su

• Use indefinite integral 2.325.6,

e ax n
(− 1) z +1
(
a z Γ − z,−ax n ) m −1
∫ x m dx = n
z=
n
with parameters [m = n, a = n = 1, z = n − 1]
to give:
(− 1)n Γ(− n + 1,− s ) sv
ea su
(− 1) ⎡Γ(− n + 1,−a − ln v ) ⎤
n
= a ⎢
e ⎣− Γ(− n + 1,−a − ln u )⎥⎦
NOTE: If
n = 2, u = 0, ln(u ) = −∞, and v = 1,
ln(v) = 0,
then Formula 4.212.3 results.
NOTE: Formula 4.212.8 can be derived by
change of variable, s = a + ln x to give,
1 es
a
(− 1)n Γ(− n + 1,−a ) ,
e a −∫∞ s n
ds =
ea
using the above indefinite integral
derivation-approach, which also results, if
v
dx
integral, ∫ , has limits set to
u (a + ln x )
n

u = 0, ln (u ) = −∞, v = 1, ln (v ) = 0.
v
dx Similar to above integral,
∫ (a − ln x )
u
n
= v
dx
∫ (a + ln x ) n

e [Γ(− n + 1, a − ln v ) − Γ(− n + 1, a − ln u )]
u
a
= (− 1) e − a [Γ(− n + 1,−a − ln v ) − Γ(− n + 1,−a − ln u )]
n

v
ln xdx • Change of variable:
∫ (a + ln x )
u
2 ds 1
s = a + ln x, = ,
dx x
av au
= − + su = ln u + a, sv = ln v + a, e s = e ln x + a
a + ln v a + ln u
(1 − a )e −a [Ei(a + lnv ) − Ei(a + lnu )] = xe ⇒ x = aa es
e
[a + ln u > 0, a + ln v > 0] I=∫
sv
x (s − a ) e (s − a )
sv s

su
s 2
ds = ∫
su
ea s 2
ds =

1 ⎡ sv e s sv s
e ⎤
⎢ ∫ ds −a ∫ 2 ds ⎥
ea⎢⎣ su s su
s ⎥⎦
• Integration by parts on first integral:

—Sect. 4.212—
 170

1
Let u = ds, du = − s −2 ds; dv = e s ds, v = e s
s
sv
es sv es
uv − ∫ vdu = +∫ dt
s su
su
s2
Then,
1 ⎡ e s sv sv e s sv s
e ⎤
I= ⎢ + ∫ 2 ds − a ∫ 2 ds ⎥ =
ea ⎢⎣ s su su s su
s ⎥⎦
⎡ e ln v+a e ln u + a
sv s
⎤
+ (1 − a )∫ 2 ds ⎥.
1 e
⎢ −
ea⎢⎣ ln v + a ln u + a su
s ⎥⎦
• Apply indefinite integral 2.325.6,
e ax
n
(− 1) z +1
(
a z Γ − z ,−ax n ) m −1
∫ x m dx = n
z=
n
with parameters [m = 2, a = n = 1, z = 1]
sv
e s ds
to give, ∫s s 2 =
u

sv
Γ(− 1,− s ) =
su
Γ(− 1,− a − ln v ) − Γ(− 1,− a − ln u )
and use Formula 8.352.3, Γ(− n, x ), to
simplify.
• Combining all calculations provides
v
ln x dx
result for ∫ .
u (a + ln x )2

NOTE: If

a = 1, and u = 1, ln(1) = 0, and v = e, ln(e) = 1, th

e
en, I = − 1 , Formula 4.212.7.
2
NOTE: If
u = 0, ln(u ) = −∞, and v = 1, ln(v) = 0,
then, Formula 4.212 5 results.
v
ln xdx • Change of variable:
∫ (a + ln x )
u
n
s = a + ln x,
ds 1
= , su =
dx x
v u
= − + ln u + a, sv = ln v + a,
(a + ln v )n −1
(a + ln u )n−1
es
Γ(− n + 1,−a − ln v ) ⎤ e =e ln x + a
= xe ⇒ x = a , ln x = s − a.
(− 1)n (n − 1 − a )e −a ⎡⎢
s a

⎥ e
⎣− Γ(− n + 1,−a − ln u )⎦

—Sect. 4.212—
 171

x(s − a ) e s (s − a )
sv sv
1
I= ∫
su
s n
ds = a
e ∫ s n ds =
su

1 ⎡ sv e s sv s
e ⎤
⎢ ∫ n−1 ds −a ∫ n ds ⎥
ea ⎢⎣ su s su
s ⎥⎦

• Integration by parts on first integral:

Let u = s1− n , du = (1 − n )s − n ds; dv = e s ds,
v = es
sv
es es
+ (n − 1) ∫
sv
uv − ∫ vdu = n −1 n
dt
s su
su s

Then,
I=
⎡ e s sv sv s sv s
⎤
⎢ n −1 + (n − 1)∫ n ds − a ∫ n ds ⎥ =
1 e e
ea ⎢⎣ s su su
s su
s ⎥⎦
⎡ e a + ln v e a + ln u ⎤
−
⎢
1 ⎢ (a + ln v )
n −1
(a + ln u )n−1 ⎥⎥
ea ⎢ sv s ⎥
⎢+ (n − a − 1)∫ n ds
e
⎥
⎢⎣ su
s ⎥⎦
sv
es
• Integral, ∫ s n ds ,solved by indefinite
su

integral 2.325.6,
e ax
n
(− 1) z +1
(
a z Γ − z ,−ax n ) m −1
∫ x m dx = n
z=
n
with parameters [m = n, a = n = 1, z = n − 1] ,
to give,
sv s
e
∫s s n ds
u

sv
= (− 1) Γ(− n + 1,− s )
n

su
n ⎡Γ (− n + 1,− a − ln v ) ⎤
= (− 1) ⎢ ⎥
⎣− Γ(− n + 1,− a − ln u )⎦
Combining all terms produces final integral
for
sv
ln x
I=∫ dx
su (a + ln x )
n

NOTE:
If

—Sect. 4.212—
 172

n = 2, u = 0, ln(u ) = −∞, and v = 1,

ln(v) = 0,
then Formula 4.212.5 results.
v v
ln xdx ln xdx
∫u (a − ln x )n Similar to above integral, ∫ (a + ln x ) n
,
u

=
u
−
v
−
sv
(s − a )e − s ds , based
∫
a
(ln u − a )n −1
(ln v − a )n−1 by solving integral, e
su
sn
(n − 1 + a )e a [Γ(− n + 1, a − ln v ) − Γ(− n + 1, a − ln u )] on change of variable, s = a − ln x.
NOTE: if
n = 2, u = 0, ln(u ) = −∞, v = 1, ln(v) = 0 ,
then Formula 4.212.6 results
v
ln xdx au av Similar to above integral,
∫ (a − ln x )
u
2
= −
ln u − a ln v − a
v
ln xdx
∫ (a + ln x ) 2

+ (a + 1)e [Ei(lnv − a ) − Ei(lnu − a )]

u
a

[ln u − a > 0, ln v − a > 0]

—Sect. 4.212—
 173

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
Formula 4.215.1 Since, by Formula 8.334.3,
π
Γ(μ )Γ(1 − μ ) =
Change Formula 4.215 1 to read: sin μπ
μ −1
1
⎛ 1⎞ π π
∫0 ⎜⎝ ln x ⎟⎠ dx = Γ(μ ) = cos ec μπ [Re μ > 0] ⇒ Γ(μ ) = =
Γ(1 − μ ) Γ(1 − μ )sinμπ
π
cosecμπ ,
Γ(1 − μ )
and it complements Formula 4.215 2
which can be re-stated,
1

∫0 ⎛ 1 ⎞ μ dx = Γ(1 − μ )
dx

⎜ ln ⎟
⎝ x⎠
as in next integral.
¾ Verified math.com

Formula 4.215.2 See explanation above for Formula

4.215.1,
μ −1
Change Formula 4.215.2 to read 1
⎛ 1⎞ π
1
π
[Re μ < 1]
∫0 ⎜⎝ ln x ⎟⎠ dx = Γ(μ ) = Γ(1 − μ ) cos ec μπ
∫0 ⎛ 1 ⎞ μ = Γ(1 − μ ) = Γ(μ ) cos ec μπ
dx

⎜ ln ⎟ ¾ Verified math.com
⎝ x⎠

n
−1 In Formula 4.215.1,
⎛ 1 ⎞2 ⎛n⎞
1

∫0 ⎜⎝ ln x ⎟⎠ dx = Γ⎜⎝ 2 ⎟⎠ [n ≥ 1] 1
⎛ 1⎞
μ −1

∫0 ⎜⎝ ln x ⎟⎠ dx = Γ(μ ), set
n n
μ −1 = −1 → μ = .
2 2
Or—Re-express integral,
1 n 1
−1

I = ∫ (− ln x ) dx = (− 1) ∫ (ln x )
n n
−1 2 −1
2 2 dx
0 0

Change of variable,
ds 1
s = ln x, = , s0 = −∞, s1 = 0,
dx x
e =e =x
s ln x

and make substitutions:

n 0 n n 0 n
−1 −1
−1 −1
I = (− 1) ∫ xs ds = (− 1) ∫s
2 2
2 2 s
e ds .
−∞ −∞
Now use indefinite integral previously

—Sect. 4.215—
 174

submitted,
(− 1)1−γ Γ(γ ,−ax n ), γ m +1
∫ x e dx = =
n
m ax
γ
,
na n
with parameters
⎡ n n⎤
⎢⎣ m = − 1 , a = n = 1, γ = to give,
2 2 ⎥⎦
⎡ ⎞0 ⎤
n n
−1 1−
⎛n
I = (− 1) ⎢(− 1) Γ⎜ ,− s ⎟
2 2

⎥
⎣ ⎝2 ⎠ − ∞⎦
⎛n ⎞ ⎛n ⎞ ⎛n⎞
= Γ⎜ ,0 ⎟ − Γ⎜ , ∞ ⎟ = Γ⎜ ⎟
⎝2 ⎠ ⎝2 ⎠ ⎝2⎠
⎛n ⎞ ⎛n⎞ ⎛n ⎞
since Γ⎜ ,0 ⎟ = Γ⎜ ⎟ and Γ⎜ , ∞ ⎟ = 0.
⎝2 ⎠ ⎝2⎠ ⎝2 ⎠
⎛n⎞
NOTE: the calculation, Γ⎜ ⎟, given in
⎝2⎠
Sect. 8.339, below.
NOTE: Formulas 4.215.3 & 4.215.4
⎛n⎞
follow from Γ⎜ ⎟ if n = 3, n = 1,
⎝2⎠
respectively.
NOTE: Set 2 → k for more general
solution, Γ(n k )
¾ Verified on math.com

1
⎛ n⎞ In Formula 4.215.1, set
∫
dx
= Γ⎜ − ⎟ [n ≥ 1, odd ] n n
⎝ 2⎠ μ −1 = − −1 → μ = − .
n
+1
0 ⎛ 1⎞ 2
⎜ ln ⎟ 2 2
⎝ x⎠ Or—Proof similar to
n
−1
⎛ 1 ⎞2 ⎛n⎞
1

∫0 ⎜⎝ ln x ⎟⎠ dx = Γ⎜⎝ 2 ⎟⎠
⎛ n⎞
NOTE: the calculation, Γ⎜ − ⎟, given in
⎝ 2⎠
Sect. 8.339, below.
NOTE: Set 2 → k for more general
solution, Γ(− n k ), n k not a natural
number.

¾ Verified on math.com
μ −1 Re-express integral,
⎛ 1⎞ ⎛ 1⎞ ⎛ 1⎞
v

∫ ⎜⎝ ln x ⎟⎠
u
dx = Γ⎜ μ , ln ⎟ − Γ⎜ μ , ln ⎟
⎝ v⎠ ⎝ u⎠
1
I = ∫ (− ln x ) dx = (− 1) ∫ (ln x ) dx .
μ −1 μ −1μ −1
1

0 0

Change of variable:

—Sect. 4.215—
 175

ds 1
s = ln x, = , su = ln u , sv = ln v,
dx x
e s = eln x = x
Then,
sv s

I = (− 1) dx = (− 1)
μ −1 μ −1

∫ xs ∫e s
μ −1 s μ −1
dx.
su su

Now use indefinite integral previously

submitted,
(− 1)1−γ Γ(γ ,−ax n ), γ = m + 1 ,
∫ =
n
m ax
x e dx γ
na n
with parameters
[m = μ − 1, a = n = 1, γ = μ ] to give,
μ −1 ⎡ sv ⎤
I = (− 1) ⎢(− 1) Γ(μ ,− s ) ⎥
1− μ

⎣ su ⎦
= Γ(μ ,− sv ) − Γ(μ ,− su )
⎛ 1⎞ ⎛ 1⎞
= Γ⎜ μ , ln ⎟ − Γ⎜ μ , ln ⎟.
⎝ v⎠ ⎝ u⎠
NOTE: Formula 4.215.1 follows if
u = 0, ln(0) = −∞, v =1, ln(1) = 0.
¾ Verified math.com
Γ(γ )
γ −1 ∞ From Form. 4.215.1,
⎛ 1⎞
( )
1
1
γ ∫⎜
ln ⎟ dt = ∫ x m exp − β x n dx = , μ −1
⎛ 1⎞
1
nβ 0 ⎝ t ⎠ nβ γ
0
∫0 ⎜⎝ ln x ⎟⎠ dx = Γ(μ ),
m +1
γ= [Re n > 0, Re β > 0, Re γ > 0] set μ = γ to get,
n
Γ(γ )
I= which is equal to
nβ γ
∞

∫x
m
(
exp− βxn dx )
0

by 3.326.2.
NOTE: see Sect. 2.71 for indefinite
integral exercise
βz 1
dt
∞
( )
exp − βx n dx β z
Γ(− z ),
Γ(γ )
n ∫⎛ 1⎞
z +1
=∫
xm
=
n
Similar to above,
nβ γ
⎜ ln ⎟
0 0
NOTE: see Sect. 2.71 for indefinite
⎝ t⎠
integral exercise
m −1
z= [Re n > 0, Re β > 0, Re z > 0]
n

—Sect. 4.215—
176
 177

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
ln (ln x )
∞ Derived in Sect. 8.367, below
∫
1
x2
dx = −C

—Sect. 4.229—
178
 179

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
1 ln( x − 1)
∞ 2 Catalan’s constant G
1 ln x
∫
4 1 x (x + 1)
dx = − ∫
4 1 x x −1
dx = G ¾ First integral,
t
∞
1 te 2
o In integral, G = ∫ t dt , use change
4 0 e +1
of variable:
set s = e t , ds = e t dt = sdt , s 0 = 1, s ∞ = ∞, ln s = t ,
to give
1
∞ ∞
1 s 2 ln s 1 ln s
I= ∫ ds = ∫ ds
4 1 s(s + 1) 4 1 s (s + 1)
¾ Second integral is derived from Formula
1
ln t
4.231.12, G = − ∫ dt ,
0
1+ t 2
with change of variable,
1 1
s = t 2 + 1, ds = 2tdt , s0 = 1, s1 = 2, t = s − 1 , ln (s − 1) 2 = ln (s − 1).
2
¾ verified math.com

—Sect. 4.241—
180
 181

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
⎛n ⎞ Re-express integral,
n Γ ⎜ + 1⎟
⎝ 2 ⎠ n(n − 2)!! π
1 n n 1
⎛ 1 ⎞ p−1
1
I = ∫ (− ln x ) x dx = (− 1) ∫ (ln x )2 x dx
2 n

∫0 ⎜⎝ ln x ⎟⎠ x dx = n +1 =
2
p −1 2
p −1

p(2 p )2 2
n
0 0
p2
[ p > 0, n = 1,3,5,...] Change of variable,
ds 1
s = ln x, = , s0 = −∞, s1 = 0,
dx x
e s = e ln x = x
and make substitutions:

( )
n n
n 0 0
n

I = (− 1) 2 ∫ s xx p −1ds = (− 1) ∫ s 2 e ds .
2 2
s p

−∞ −∞
Now use indefinite integral previously
submitted,
(− 1)1−γ Γ(γ ,−ax n ), γ = m + 1 ,
∫ =
n
m ax
x e dx γ
na n
with parameters
⎡ n n ⎤
⎢⎣ m = , a = p , n = 1, γ = + 1 to give,
2 2 ⎥⎦
n ⎡ −
n ⎤
⎛n ⎞0 ⎥
(− 1) 2 ⎢⎢(− 1) 2
Γ⎜⎜ + 1,− ps ⎟⎟ ⎥
⎢ ⎝2 ⎠ − ∞⎥
I= ⎣ ⎦
n
+1
p2
⎛n ⎞ ⎛n ⎞ ⎛n ⎞
Γ⎜⎜ + 1,0 ⎟⎟ − Γ⎜⎜ + 1, ∞ ⎟⎟ Γ⎜⎜ + 1⎟⎟
= ⎝ 2 ⎠ ⎝ 2 ⎠ = ⎝ 2 ⎠ =
n n
+1 +1
p2 p2
⎛n⎞
Γ⎜⎜ ⎟⎟
n ⎝ 2 ⎠ = n(n − 2 )!! π
2 n n 2
+1 p (2 p ) 2
p2

⎛n ⎞ ⎛n ⎞
by noting Γ⎜ + 1,0 ⎟ = Γ⎜ + 1⎟ and
⎝2 ⎠ ⎝2 ⎠
⎛n ⎞
Γ⎜ + 1, ∞ ⎟ = 0, and, by formula, Sect. 8.339,
⎝2 ⎠
below,
⎛n⎞ π (n − 2 )!!
Γ⎜ ⎟= n −1
[n = 1,3,5,K]
⎝ ⎠
2 2 2

π π
NOTE: if n = 1, I = 3
=
2p 2 2 p3

3 π
of integral 4.269.3. If n = 3, I =
4 p5
—Sect. 4.269—
 182

¾ Verified on math.com
⎛ n⎞ Re-express integral,
Γ⎜1 − ⎟
n+3

2 ⎠ (− 1) (2 p )2
n
p −1
π x p −1 x p −1
1 2 1 1
x ⎝ I =∫
1
n ∫
∫0 n
dx = n
=
p(n − 2)!!
n dx =
0 (ln x ) 2
n
dx
2
(− ln x ) (− 1)
1−
⎛ 1 ⎞2
2 2
0
p 2
⎜ ln ⎟
⎝ x⎠ −
n 1
= (− 1) ∫ (ln x )
n
−
[ p > 0, n = 1,3,5,...] x p −1dx
2
2
0
Change of variable,
ds 1
s = ln x, = , s0 = −∞, s1 = 0,
dx x
e s = e ln x = x
and make substitutions:

∫ s (e ) ds .
n
n 0 0 n
n −
− −
I = (− 1) ∫s ds = (− 1)
− 2 p −1 2
2 s p
2 xx
−∞ −∞
Now use indefinite integral 2.325.6,
n (− 1)z +1 a z Γ⎛⎜ − z,−ax n ⎞⎟
∫
e ax ⎝ ⎠ m −1
dx = z=
xm n n

⎡ n n ⎤
with parameters ⎢m = , a = p, n = 1, z = − 1⎥
⎣ 2 2 ⎦
to give,
2 ⎡ 0 ⎤
n n

(− 1) ⎢(− 1) Γ⎛⎜1 − n ,− ps ⎞⎟ ⎥
−
2

⎣ ⎝ 2 ⎠ − ∞⎦
I= n
1−
p 2
⎛ n ⎞ ⎛ n ⎞ ⎛ n⎞
Γ⎜1 − ,0 ⎟ − Γ⎜1 − , ∞ ⎟ Γ⎜1 − ⎟
= ⎝
2 ⎠ ⎝ 2 ⎠ = ⎝ 2⎠
n n
1− 1−
p 2 p 2
⎛ n ⎞ ⎛ n⎞
by noting Γ⎜1 − ,0 ⎟ = Γ⎜1 − ⎟ and
⎝ 2 ⎠ ⎝ 2⎠
⎛ n ⎞
Γ⎜1 − , ∞ ⎟ = 0.
⎝ 2 ⎠
⎛ n⎞
Γ⎜ 1 − ⎟
⎝ 2 ⎠ can be reduced by noting that
n
1−
p 2
⎛ n⎞ n ⎛ n⎞
Γ⎜1 − ⎟ = − Γ⎜ − ⎟ by Formula 8.331, and
⎝ 2⎠ 2 ⎝ 2⎠
by formula, Sect. 8.339, below,
n +1
n +1
n −1 ⎟
⎛ ⎞
(−1) 2
2n π ⎜⎜ ⎟!
Γ⎛⎜⎜ − n ⎞⎟⎟ = ⎝ 2 ⎠
=
(− 2) 2
π
[n = 1,3,5,K]
⎝ 2⎠ n! (n)!!
Thus, after simplifying,

—Sect. 4.269—
 183

⎛ n⎞
Γ⎜1 − ⎟
n+3

2 ⎠ (− 1) (2 p )2
n
p −1
π
1 2
x ⎝
∫0 n
dx =
1−
n
=
p (n − 2)!! 2
⎛ 1 ⎞ 2
p 2
⎜ ln ⎟
⎝ x⎠
π
NOTE: if n = 1, I = of integral 4.269.4.
p
If n = 3, I = −2 pπ .

¾ Verified on math.com

—Sect. 4.269—
 184

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
Γ(n + 1) Follows from Formula 4.272.6,
1

∫ (ln x )
n
x p dx = (− 1)
n
[Re n + 1 > 0] μ −1
Γ (μ )
∞

( p + 1)n+1 ⎛ 1⎞
1

0 ∫ ⎜⎝ ln x ⎟⎠
0
xν −1dx =
ν μ
= ∫ x μ −1e −ν x dx,
0

by substituting ν − 1 → p, μ − 1 → n ,
1
setting ln x = − , and factoring out
ln x
(− 1)n ,
giving,
n Γ(n + 1)
n
⎛ 1⎞
1
I = (− 1) ∫ ⎜ ln ⎟ x p dx = (− 1)
n

0⎝
x⎠ ( p + 1)n +1
∫
1
NOTE: following f ( x)dx integrals of
0

similar form are derivable from Formula

4.272.6, ⎛ ln 1 ⎞ μ xν dx = Γ(μ ), , by
1 −1

∫ ⎜⎝ x ⎟⎠
−1

νμ 0

substituting appropriate parameters for

μ ,ν .
¾ Verified on math.com
1 Follows from above integral,
∫ (ln x )x
1
p
dx = − n Γ(n + 1)
( p + 1)2
1

∫0 (ln x ) x dx = (− 1) ( p + 1)n+1 ,
n p with
0

n = 1, Γ(2 ) = 1.
¾ Verified on math.com
1 In Formula 4.272.6, referenced above,
1
⎛ 1 ⎞ 2 ν −1
−n
(− 1) n +1
(2n − 1)(2ν )n−1 π
∫0 ⎜⎝ ln x ⎟⎠ x dx = (2n − 1)!! ν
1
set μ − 1 = − n , to get
2
[Reν > 0] ⎛ 3 ⎞
Γ⎜ − n ⎟
⎝2 ⎠ , and simplfy.
3
−n
ν 2

Or—Re-express integral,
1 1
−n

I = ∫ (− ln x ) xν −1 dx =
2

0
1 1
−n

(− 1) ∫ (ln x ) 2 −n xν −1dx
2
1

0
Change of variable,
ds 1
s = ln x, = , s0 = −∞, s1 = 0,
dx x
e s = e ln x = x
and make substitutions:

—Sect. 4.272—
 185

0 1
−n

I = (− 1)
1

∫s
−n
xxν −1 ds =
2
2
−∞
1
.
(e ) ds
1 1 −n
−n

(− 1) ∫ s
2
2
s v

Now use indefinite integral previously

submitted,
(− 1)1−γ Γ(γ ,−ax n ), γ = m + 1 ,
∫ =
n
m ax
x e dx γ
na n
with parameters
⎡ 1 3 ⎤
⎢⎣m = 2 − n, a = ν , n = 1, γ = 2 − n ⎥⎦ to give,
⎡ 0 ⎤
1 1

(− 1) ⎢(− 1) Γ⎛⎜ 3 − n,−νs ⎞⎟ ⎥

−n − +n
2 2

⎣ ⎝2 ⎠ − ∞⎦
I= 3
−n
ν2
⎛3 ⎞ ⎛3 ⎞ ⎛3 ⎞
Γ⎜ − n,0 ⎟ − Γ⎜ − n, ∞ ⎟ Γ⎜ − n ⎟
= ⎝ ⎠ ⎝2 ⎠= ⎝2 ⎠
2
3 3
−n −n
ν 2
ν 2

⎛3 ⎞ ⎛3 ⎞
since Γ⎜ − n,0 ⎟ = Γ⎜ − n ⎟ and
⎝2 ⎠ ⎝2 ⎠
⎛3 ⎞
Γ⎜ − n, ∞ ⎟ = 0.
⎝2 ⎠
Now, rearrange numerator,
⎛3 ⎞ ⎡⎛ 1 ⎞ ⎤
Γ⎜ − n ⎟ = Γ ⎢⎜ − n ⎟ + 1⎥ =
⎝2 ⎠ ⎣⎝ 2 ⎠ ⎦
⎛1 ⎞ ⎛1 ⎞
⎜ − n ⎟Γ⎜ − n ⎟,
⎝2 ⎠ ⎝2 ⎠
where,
⎛1 ⎞ (− 1) π 2
n n
Γ⎜ − n ⎟ = by Formula 8.339 .3
⎝2 ⎠ (2n − 1)!!
Thus,

I=
(− 1) (1 − 2n ) π 2 n −1 n − 2
n
ν
3

(2n − 1)!!
=
(− 1)n+1 (2n − 1)(2ν )n−1 π
(2n − 1)!! ν

NOTE: unreduced expression,

(2n −1) = 1
(2n −1)!! (2n - 3)!!
allows n = 0 calculation.
—Sect. 4.272—
 186

¾ Verified on math.com
1 Proof similar to above integral,
1
⎛ 1⎞
n+
2 (2n + 1)(2n − 1)!! π
[Reν > 0] [
∫ ⎜⎝ ln x ⎟⎠ xν −1dx = 1

(2ν )n+1 ν
1 −n
⎛ 1 ⎞ 2 ν −1 (−1) (2n −1)(2ν ) n +1 n −1
π
0
∫0 ⎜⎝ ln x ⎟⎠ x dx = (2n −1)!! ν
[Reν > 0]
¾ Verified on math.com

∞
(ln x ) p ± a ± n−1 dx = Γ( p ± a ± n ) [Re p ± a ± n > 0] s = ln x,
ds 1
= , s1 = 0, s∞ = ∞, e s = e ln x = x
∫
1
2
x dx x

giving,
∞

∫ x(s )
p ± a ± n −1
ds ∞
= ∫s
p ± a ± n −1
I= 0
2
e − s ds =
x 0

Γ( p ± a ± n )
by Formula 3.326.2 with parameters
[m = p ± a ± n − 1, β = n = 1, γ = p ± a ± n]
Verified on math.com
∞
(ln x ) p + a −1 dx = Γ( p + a ) [Re p + a > 0] Change of variable,
∫
1 x2 ds 1
s = ln x, = , s1 = 0, s∞ = ∞, e s = eln x = x
dx x
giving,

∫ x (s )
p + a −1
ds ∞
e −s ds = Γ( p + a )
= ∫s
p + a −1
I= 0
2
x 0

by Formula 3.326.2 with parameters

[m = p + a − 1, β = n = 1, γ = p + a ]
NOTE: in above integral, Γ( p ± a ± n ), set
parameter n = 0.
¾ Verified on math.com
∞
(ln x ) p−1 dx = Γ( p ) [Re p > 0] Set a = 0 in above integral,
∫ 2 ∞
(ln x ) p+a−1 dx = Γ( p + a )
1
x
∫1 x 2
NOTE: alternative expression for Formula
4.215.5,
Γ(μ ).
NOTE: in above integral, Γ( p ± a ± n ), set
parameters n ± a = 0.
¾ Verified math.com

—Sect. 4.272—
 187

v
(ln x )p −1 dx = Γ( p, ln u ) − Γ( p, ln v ) Change of variable:
∫
u
x2 ds 1
s = ln x, = , su = ln u, sv = ln v,
dx x
e =e =x
s ln x

and make substitutions:

sv p −1 s
v p −1
s
I = ∫ 2 xds = ∫ s e − s ds .
su
x su

Now use indefinite integral 2.33.10,

Γ(γ , βx n ) m +1
∫
m − βx
= − ,γ =
n
x e dx γ
,
na n
with parameters
[m = p − 1, a = ν , n = 1, γ = p ] to give,
I = −[Γ( p, sv ) − Γ( p, su )]
= Γ( p, ln u ) − Γ( p, ln v )
NOTE: if u = 1, v = ∞,

I =∫
∞
(ln x )p −1 dx = Γ( p ) .
1
x2
∞ Change of variable:
dx
∫ (ln x ) = Γ(1 − p ) [Re p < 1] ds 1
s = ln x, = , s1 = 0, s∞ = ∞,
p
1 x2
dx x
e s = e ln x = x
and make substitutions:
sv sv − s
∞
I = ∫ p 2 xds = ∫ p ds = −Γ(1 − p, s )
x e
su s x su
s 0
= Γ(1 − p,0 ) − Γ(1 − p, ∞ ) = Γ(1 − p )
by indefinite integral 2.33.17,
e − βx
n
(
β z Γ − z, βx n )m −1
∫ xm dx = −
n
,z =
n
with parameters
[m = p, β = n = 1, z = p − 1]
and,
Γ(a,0 ) = Γ(a ), Γ(a, ∞ ) = 0.
¾ Verified math.com

Γ(1 − p ) • First form is similar to above integral,

∞

∫ (ln x )
dx
p +1
= Γ(− p ) = − [p ∉ N ] set p → p + 1.
1 x2 p
• Second form by rearrangement of
Formula 8.331,
Γ (x + 1)
• Γ (x ) = with x set to
x
− x; e.g.,

—Sect. 4.272—
 188

⎛ 1⎞
Γ⎜ − ⎟ = −2 π .
⎝ 2⎠
• NOTE: p not a natural number N
( p ∉ N ) is understood for complete
gamma function, Γ( x ), x < 0.
¾ Verified math.com

—Sect. 4.272—
 189

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
1−
n
⎛ n⎞ • Change of variable,
Γ⎜ 1 − ⎟
1 2
q q s = ln x + 1
⎝ 2⎠
e
x
∫ n dx = q ds 1
= , s0 = −∞, s1 =
x[− (1 + ln x )]
0 2 e
dx x e

[q > 0, n odd ] ⎛1⎞

ln⎜ ⎟ + 1 = −1 + 1 = 0
⎝e⎠
s
s 1 q
e e
e s = xe, x = ⇒x = 1 q

e
eq
• Re-express integral,
1 1
−1
e q
1 x
I= ∫
(− 1) (ln x + 1)2
n n
2 0

• Substitute,
1
−1
n 0 xq x
I = (− 1) ∫
−
2
n
=
−∞ 2
s
(− 1)
n
− 0 s n
−
∫e
2 q 2
1
s .
q −∞
e
• Use indefinite integral previously
submitted,
∫ x e dx =
m ax n

(− 1)1− γ (
Γ γ ,− ax n
,γ =
m +1 )
γ
na n
with parameters
⎡ n 1 n⎤
⎢m = − 2 , a = q , n = 1, γ = 1 − 2 ⎥ to
⎣ ⎦
give,

—Sect. 4.274—
 190

⎡ ⎤
⎢ ⎛ n s⎞ 0
⎥
Γ⎜⎜1 − ,− ⎟⎟
n⎢ ⎥
I = (− 1) 2 ⎢(− 1) 2 ⎝
−
n 2 q⎠
n
⎥
⎢ 1
⎛ 1 ⎞
1−
2 ⎥
⎢ e q ⎜⎜ ⎟⎟ −∞ ⎥
⎢⎣ ⎝q⎠ ⎥⎦
⎛ n ⎞ ⎛ n ⎞
Γ⎜1 − ,0 ⎟ − Γ⎜1 − , ∞ ⎟
= ⎝
2 ⎠ ⎝ 2 ⎠=
1
n
e q ( q −1 )1− 2
⎛ n⎞
n
1−
q Γ⎜1 − ⎟
2

⎝ 2⎠.
q
e
NOTE: Formula 2.274 results if
n = 1.

—Sect. 4.274—
191
 192

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
∞ Euler’s constant, Derived in Sect. 8.367, below
⎡ 1 1 ⎤ dx
∫ ⎢⎣ x − 1 − x ln x ⎥⎦
1
x
=C

• Change of variable,
∞
ds 1
∫ x (ln p + ln x ) = − pli(− p )
dx s = ln p + ln x, = , s1 = ln p + 0 = ln p, s∞ = ∞,
2
1
dx x
es
e s = e ln p e ln x = px → x =
p
• Then,
∞ ∞ ∞
xds ds e − s ds
I = ∫ 2 ds = ∫ =p∫
ln p
x s ln p e
s ln p
s
s
p
∞
= pEi(− s ) = p[Ei(− ∞ ) − Ei(− ln p )] = − pli(− p )
ln p
¾ Not verifiable on math.com; nor was Formula 4.281.2
x n + a −1 • Change of variable:
( )
u

∫0 ln x dx = li u
n+a
[u > 1] dx
s = ln x, ds = , s 0 = −∞, su = ln u, e s = x, x n + a = e (n + a )s , to
x
give
ln u n + a −1 ln u ( n + a )s
ds = Ei[(n + a ) ln u ] − 0 =
x e
I= ∫ xds = ∫
−∞
s −∞
s
( ) ( )
li e (n + a ) ln u = li u n + a
by relation, e x ln a = a x , and Ei(− ∞ ) = 0.
¾ Verified math.com
x n −1 x n + a −1
( ) ( )
u u

∫0 ln x dx = li u
n
[u > 1] In above integral, ∫ dx = li u n + a , set a = 0.
0
ln x
¾ Verified math.com

—Sect. 4.281—
 193

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
∞
⎡ 1 1 ⎤ dx In Formula 8.367.9,
∫ ⎢⎣ x ln x − ln x(1 + ln x )⎥⎦ = −C ∞
⎡ 1 ⎤ dt
∞
⎡1 1 ⎤ dt
x C = − ∫ ⎢e −t − ⎥ = − ∫ ⎢ − ,
e 1 + t ⎥⎦ t
1
0 ⎣
1+ t ⎦ t 0 ⎣
t

set s = e t , ds = e t dt = sdt , ln s = t , s0 = 1, s∞ = ∞ , and

substitute to get
∞
⎡ 1 1 ⎤ ds
C = −∫ ⎢ −
1 ⎣
s ln s ln s(1 + ln s ) ⎥⎦ s
¾ Verified math.com
1
⎡1 1 ⎤ • From integral, Sect. 8.367, below,
∫ ⎢⎣ x + ln(1 − x )⎥⎦ dx = C 0
⎡ et et ⎤
0 C = ∫⎢ − t ⎥ dt .
−∞ ⎣
t e − 1⎦
⎡1 1 ⎤
1
• Substitute into hypothesis C = ∫ ⎢ +
t ln (1 − t ) ⎥⎦
dt ,
0⎣

s = ln (1 − t ), ds = −
1
dt , e s = e ln(1−t ) = 1 − t ,
1− t to get,
s 0 = 0, s1 = −∞,
0
⎡ et et ⎤
C= ∫ ⎢ t et − 1⎥⎦ dt .
−∞ ⎣
−

¾ Verified math.com

—Sect. 4.283—
194
 195

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
• Change of variable:
[ln a − e ]
1

∫ ln(a + ln x )x Ei(aμ )
μ −1 1 −μ a
dx = dx
0
μ s = a + ln x, ds = , e s = e a + ln x = e a x,
[Re μ > 0, a > 0] x
x = e e ⇒ x = e μs e − μa
s −a μ

s0 = a + (−∞) = −∞
s1 = a + 0 = a
a a

∫ x ln xx dx = e ∫ e ln sds
μ −1 − μa μs
I=
−∞ −∞

• Integration by parts:
ds
u = ln s, du =
s
1
dv = e μs ds, v = ∫ e μs ds = e μs
μ
1 1 e μs
uv − ∫ vds =
μ∫ s
e μs ln s − ds
μ
s=a s=a
− Ei(μs )
1 1
= e μs ln s
μ s = −∞ μ s = −∞

Ei(μa )
1 1
= e μa ln a −
μ μ

• Combining results:
⎡1 ⎤
I = e −μa ⎢ e μa ln a − Ei(μa )⎥
1
⎣μ μ ⎦
[
= ln a − e −μa Ei(aμ )
1
μ
]
¾ Math.com unable to do but returned indef.
solution:

x m ln (a − ln x ) 1 −μa
− e Ei[(μ )(a + ln x )]
m μ

—Sect. 4.326—
196
 197

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
∞
• Γ (1) (1) = C , Euler’s constant
(− 1) ⎛⎜⎜ ⎞⎟⎟(ln μ ) Γ( ) (1)
n n
∫ e (ln x ) dx =
1
− μx
∑ n−k
n k k

μ k =0 k ⎝ ⎠ • Derived in V.H. Moll, “The integrals of

0

[Re μ > 0, n = 0,1,2,...]

Gradshteyn and Rhyzik. Part 4. The Gamma
Function. Preliminary Version. Last Update:
September 7, 2006”, page 3, Formula 2.17.
¾ NOTE: GR Formulas 4.331.1,
4.335.1 , 4.335.3, follow for n = 1,2,3,
respectively.
v
• In integral solution below, Sect. 4.337,
∫e
−μ x
ln xdx = v
u
∫e
−μ x
ln (a + bx )dx =
1
μ
[e −μ u
ln u − e −μ v ln v + Ei(− μ v ) − Ei(− μ u ) ] u
⎧e − μ u ln (a + bu ) − e − μ v ln (a + bv ) + ⎫
1 ⎪ μa ⎪
[Re μ > 0] ⎨ b ⎡ ⎛
⎩ ⎣ ⎝
μa ⎞
μ ⎪e ⎢Ei⎜ − μ v − ⎟ − Ei⎛⎜ − μ u − ⎞⎟⎥ ⎪
b ⎠ ⎝
μ
b
a ⎤ ⎬,
⎠⎦ ⎭
set a = 0, b = 1
• Or—integration by parts:
u = ln x, dv = e − μx dx
NOTE: math.com provides following
indefinite solution:
Ei(− μ x ) − e − μx ln( x)
∫
− μx
e ln xdx = .
μ
v Similar to above integral,
∫ e ln xdx =
μx
v
u
∫e
−μ x
ln ( x )dx =
1
[e μv
ln v − e μ u ln u − Ei(− μ v ) + Ei(− μ u )] u

μ 1
[e −μ u
ln u − e −μ v ln v + Ei(− μ v ) − Ei(− μ u ) ]
[μ ≠ 0] μ

Or—Integration by parts:
u = ln x, dv = e μx dx
NOTE: math.com provides following
indefinite integral:
e μx ln x − Ei(μx )
∫
μx
e ln xdx = .
μ

—Sect. 4.331—
 198

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
v
• Integration by parts:
∫e
−μ x
ln (a + bx )dx =
u = ln (a + bx ), du =
b
u dx
a + bx
⎧e − μ u ln (a + bu ) − e − μ v ln (a + bv ) + ⎫
1 ⎪ μa ⎪ 1
dv = e − μ x dx, v = ∫ e − μ x dx = − e − μ x
⎨ b ⎡ ⎛ μa ⎞ ⎛ μa ⎞⎤ ⎬ μ
μ ⎪e ⎢Ei⎜ − μ v − ⎟ − Ei⎜ − μ u − ⎟⎥ ⎪
⎩ ⎣ ⎝ b ⎠ ⎝ b ⎠⎦ ⎭ I=
⎡ a a ⎤ ln (a + bx ) − μ x b
e −μ x
⎢ arg b + u < π , arg b + v < π , Re μ > 0⎥ uv − ∫ vdu = −
μ ∫ a + bx
e +
dx
⎣ ⎦ μ
• Apply integration limits, u, v :
ln (a + bx )e − μ x v b e − μ x
v
I =− + ∫ dx
μ u μ u a + bx
e − μ u ln (a + bu ) − e − μ v ln (a + bv ) e −μ x
v
b
μ ∫u a + bx
= + dx
μ
• Solve integral:
e−μ x
∫ a + bx dx is solved with integral solution in
Sect 3.35, above,
e − ρx
v

∫ (ax ± b )
u
n
dx =

ρb
±
ρ n −1e a
⎡ ⎛ ρb ⎞ ⎛ ρb ⎞⎤
⎢Γ⎜ − n + 1, ρu ± a ⎟ − Γ⎜ − n + 1, ρv ± a ⎟⎥
⎝ ⎠ ⎝ ⎠⎦
n
a ⎣
⎡ ⎛b ⎞ ⎛b ⎞ ⎤
⎢ ρ > 0, arg⎜ ± u ⎟ < π , arg⎜ ± v ⎟ < π ⎥ ,
⎣ ⎝a ⎠ ⎝a ⎠ ⎦
with parameters [ρ = μ , b = a, a = b, n = 1] giving:
e−μ x
∫ a + bx dx =
μa
e b
⎡ ⎛ μ b ⎞⎤ ⎡ ⎛ μ b ⎞⎤
⎢ Γ⎜ 0, μ u + ⎟⎥ − ⎢Γ⎜ 0, μ v + ⎟
n ⎣ ⎝ a ⎠⎦ ⎣ ⎝ a ⎠⎥⎦
μa
e ⎡ ⎛b
μb⎞ ⎛ μ b ⎞⎤
= ⎢ − Ei⎜ − μ u − ⎟ + Ei⎜ − μ v − ⎟
n ⎣ ⎝ a ⎠ ⎝ a ⎠⎥⎦
by relation, Γ(0, x ) = − Ei(− x ) , Formula 8.359.1.

• Combine results:

—Sect. 4.337—
 199

e − μ u ln (a + bu ) − e − μ v ln (a + bv )
I=
μ
⎧ μa ⎫
b ⎪e b ⎡ ⎛ μ b ⎞⎤ ⎛ μ b ⎞⎪
+ ⎨ Ei⎜ − μ v − ⎟ − Ei⎜ − μ u − ⎟⎬
μ ⎪ b ⎢⎣ ⎝ a ⎠⎥⎦ ⎝ a ⎠⎪
⎩ ⎭
⎧e ln (a + bu ) − e ln (a + bv ) +
−μ u −μ v
⎫
⎪
1 ⎪ μa ⎪⎪
= ⎨ b ⎡
μ ⎪e ⎛ μ b ⎞⎤ ⎛ μ b ⎞⎬
⎢ Ei⎜ − μ v − ⎟⎥ − Ei⎜ − μ u − ⎟⎪
⎪⎩ b ⎣ ⎝ a ⎠⎦ ⎝ a ⎠⎪⎭
NOTE: Mathematica produced following
indefinite integral solution:
∫ e ln(a + bx )dx
−μ x

μa
⎛ μa ⎞ − μ x
e b Ei⎜ − xμ − ⎟ − e ln (a + bx )
= ⎝ b ⎠
μ
v Similar to above integral,
∫ e ln(a − bx )dx =
−μ x
v
u
∫e
−μ x
ln (a + bx )dx
⎧e − μ u ln (a − bu ) − e − μ v ln (a − bv ) + ⎫ u
1 ⎪ μa ⎪
⎨ − b ⎡ ⎛ μa ⎞ ⎛ μa ⎞⎤ ⎬
μ ⎪e ⎢Ei⎜ − μ v ⎟ − Ei⎜ − μ u ⎟⎥ ⎪
⎩ ⎣ ⎝ b ⎠ ⎝ b ⎠⎦ ⎭
⎡ a a ⎤
⎢ arg b − u < π , arg b − v < π , Re μ > 0⎥
⎣ ⎦

μa
1⎡ ⎛ μa ⎞⎤
∞ v

∫0 e ln(a + bx )dx = μ ⎢ln a − e b Ei⎜⎝ − b ⎟⎠⎥

−μ x
In above integral, ∫ e −μ x ln (a + bx )dx =
⎣ ⎦ u

⎡ a ⎤ ⎧e ln (a + bu ) − e − μ v ln (a + bv )
−μ u
⎫
⎢ arg b < π , Re μ > 0⎥ 1 ⎪ μa ⎪
⎣ ⎦ ⎨ ⎡ ⎛ μa ⎞ ⎛ μa ⎞⎤ ⎬ ,
μ ⎪+ e ⎢Ei⎜ − μ v − ⎟ − Ei⎜ − μ u − ⎟⎥ ⎪
b

⎩ ⎣ ⎝ b ⎠ ⎝ b ⎠⎦ ⎭
set u = 0, v = ∞ and note Ei(− ∞ ) = 0.
NOTE:
If b = 1, a = β , Formula 4.337.1 results.
If a = 1, b = β , Formula 4.337.2 results.

∞ v

∫e
−μ x
ln (a + bx )dx = In above integral, ∫ e − μ x ln (a + bx )dx =
u u

1 ⎡ −μ u
μa
⎛ μ a ⎞⎤ ⎧e ln (a + bu ) − e − μ v ln (a + bv )
−μ u
⎫
⎢e ln (a + bu ) − e Ei⎜ − μ u − ⎟⎥ 1 ⎪ μa ⎪
b
μ⎣ ⎝ b ⎠⎦ ⎨ ⎡ ⎛ μa ⎞ ⎛ μa ⎞⎤ ⎬ ,
μ ⎪+ e ⎢Ei⎜ − μ v − ⎟ − Ei⎜ − μ u − ⎟⎥ ⎪
b

⎩ ⎣ ⎝ b ⎠ ⎝ b ⎠⎦ ⎭
set u = u, v = ∞ and note Ei(− ∞ ) = 0.
—Sect. 4.337—
 200

⎡ a ⎤
⎢ arg b + u < π , Re μ > 0⎥
⎣ ⎦

μa
1⎡ ⎛ μa ⎞⎤
∞ Similar to above integral,
−
∫0 e ln(a − bx )dx = μ ⎢ln a − e b Ei⎜⎝ b ⎟⎠⎥
−μ x
∞

⎣ ⎦ ∫e
−μ x
ln (a + bx )dx above
⎡ a ⎤ 0

⎢ arg b < π , Re μ > 0⎥

⎣ ⎦
∞ Similar to above integral,
∫ e ln(a − bx )dx =
−μ x
∞
u
∫e
−μ x
ln (a + bx )dx above
μa
1 ⎡ −μ u − ⎛μa ⎞⎤ u

⎢e ln (a − bu ) − e Ei⎜ − μ u ⎟⎥
b
μ⎣ ⎝ b ⎠⎦
⎡ a ⎤
⎢ arg b − u < π , Re μ > 0⎥
⎣ ⎦

—Sect. 4.337—
201
202
 203

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
Either add following to Sect. 8.212 or From definition,
Sect. 8.359 ∞

∫t
α −1 −t
e dt = Γ(α , x ) ⇒ ±Γ(0,± x ) =
− Γ(0, x ) = Ei(− x ) x
∞
− Γ(0,− x ) = Ei( x ) e −t
± ∫ dt = m Ei(m x ), x > 0
Γ(0,− x ) = −Ei( x ) ±x
t
or replace Formula 8.359 1, NOTE: Ei(0) is not defined.
Γ(0, x ) = −Ei(− x ) , with ¾ Verified math.com
± Γ(0,± x ) = mEi(m x ) .
(
Ei ± x n = ) •
Solve separately for ± forms. Hypothesis, first
∞ ∞ form: in Formula 8.212 5,
e − xt e −t
± e±x ∫0 x n−1 m t ∫0 x n m t dt
± xn
= ±
n
∞
dt e e − xt
Ei(± xy ) = ± e ± xy ∫ dt , set
m
[x > 0, n ≥ 1] 0
y t
(
y = x n−1 ⇒ Ei ± xx n−1 = Ei ± x n . ) ( )
∞ − xt
Then, Ei(x n ) = e x
e
∫x
n

n −1
dt .
0
−t
Proof—
In integral Sect. 3.462, above,
ρb
∞ −
− ρx
n ρ ρb ⎞
n −1 a
⎛
∫0 (b − ax )n dx = (− 1) a n
e e
Γ⎜ − n + 1, − ⎟ , set
⎝ a ⎠
ρ = x, a = n = 1, x = t , b = x n−1 ,
to give,
∞
e − xt
∫0 x n−1 − t dt = −e Γ(0, − x ) = e Ei(x ) , so that
− xn n −xn n

( )
e x e − x Ei x n = Ei x n
n n
( )
∞
e −t
o Second form , Ei( x ) = e x ∫0 x n − t dt , follows
n

using same approach, switching roles of x in

numerator & denominator, and applying
ρb
∞ −
− ρx
n ρ ρb ⎞
n −1 a
⎛
∫0 (b − ax )n dx = (− 1) a n
e e
Γ⎜ − n + 1, − ⎟ with
⎝ a ⎠
ρ = 1, a = n = 1, x = t , b = x n .
NOTE: Similar logic applies to derivation of Formula
8.212 5, Ei(+ xy ), where

—Sect. 8.212—
 204

∞ ∞
e − xt e − t dt
∫ dt = ∫ = −e − xy Γ(0,− xy ),
0
y −t 0
xy − t
[
so that Ei(xy ) = e xy − e − xy Γ(0,− xy ) = Ei( xy). ]
This result seems consistent with the definition of
∞ −t
Ei(x ) = − ∫ dt = −Γ(0,− x ) in Formula 8.211.1 if xy
e
−x
t
serves the role of x.
• Proof is similar for
∞ ∞
e − xt e −t
( n
)
Ei − x n = −e − x ∫ n −1 dt = −e − x ∫ n
n
dt .
0 x +t 0 x +t

Apply derived Formula, 3.462 15,

ρb
∞
e − ρx ρ n−1e a ⎡ ⎛ ρb ⎞⎤
∫0 (ax + b)n dx =
a n ⎢ Γ⎜ − n + 1, ⎟⎥
⎣ ⎝ a ⎠⎦
with parameters,
ρ = x(orρ = 1), a = n = 1, x = t , b = x n−1 or b = x n , to give ( )
∞ − xt ∞ −t

∫x
e
n −1
+t
dt = ∫
e
x +t
n
(
dt = e x Γ 0, x n = −e x Ei − x n ,
n
) n
( )
[ ]
0 0

so that, Ei(− x n ) = − e − x − e x Ei(− x n ) = Ei(− x n ).

n n

∞ ∞
e − xt e −t In above,
Ei(± x ) = ± e ± x ∫ dt = ±e ± x ∫ dt ∞ ∞
e − xt e −t
1m t xmt Ei(± x n ) = ± e ± x ∫ n −1 dt = ± e ± x ∫ n
n n
0 0 dt
[x > 0] 0 x m t 0 x m t
set n = 1.
∞ ∞
⎛ 1 ⎞ e − xt e−x
Ei⎜ ± n ⎟ = In above, Ei(± x n
) = ±e ± xn
∫0 x n−1 m t dt = ±e ∫0 x n m t dt ,
± xn

⎝ x ⎠
t
1 ∞ − 1 ∞ 1
± e x ± e −t set x = .
±e xn
∫ 1
0 n −1 mt
dt = ±e xn
∫
0
1
m t
dt x
x n
x

[x > 0, n ≥ 1]
⎛ 1⎞ −
t
Ei⎜ ± ⎟ = ⎛ 1 ⎞ ±
1 ∞
e x ±
1 ∞
e −t
⎝ x⎠ In above, Ei⎜ ± n ⎟ = ±e
⎝ x ⎠
xn
∫ 1
mt
dt = ±e xn
∫ 1
mt
dt ,
t 0 n −1 0
1∞ − 1∞ −t
x x
± x ±
e e set n = 1.
±e x
∫ 1 m t dt = ±e ∫
0
x
1
0 x mt
dt

[x > 0]
Ei(x ± y ) = • First form hypothesis: in Formula 8.212.2,
⎡ 1 ∞
e −t ⎤ ⎡ ∞
e − t dt ⎤
x± y
⎢ +∫ ⎥ = Ei( x ) = e ⎢ + ∫
x 1
2⎥
, set x = x + y . Then
⎣ x 0 (x − t ) ⎦
e dt
⎣ x ± y 0 (y ± x m t)
2
⎦
∞
e −( x ± y )t e −t
∞ the integral in proposed formula Ei( x + y ) is
e x± y ∫ dt = e x ± y ∫ dt solved by integral in Sect. 3.462, above,
0
1 − t 0
x ± y − t
—Sect. 8.212—
 205

[ y > 0, x > 0] ∞
e − ρx
∫ (b − ax )
0
n
dx =

ρb
−

(− 1)n ρ ρb ⎞
n −1
e a ⎛
Γ⎜ − n + 1, − ⎟
⎝ a ⎠
n
a
by setting ρ = 1, b = x + y, a = 1, x = t , n = 2,
giving
∞
e −t
∫0 ( y + x − t )2 dt =
(− 1)2 e −( x + y ) {Γ[− 1,−(x + y )]} =
⎡ e x+ y ⎤
⎢ Ei( x + y ) −
−( x + y )
e ⎥
⎣ x + y⎦
by Formula 8.352 3, Γ(− n, x ) . Substitution into
⎡ 1 ∞
e −t ⎤
Ei(x + y ) = e ⎢ +∫ x+ y
dt ⎥ leads to result.
⎣ x + y 0 (y + x − t) ⎦
2

Ei( x − y ) is similar to Ei( x + y ) , except the integral,

∞
e −t
∫ (y − x + t)
0
2
dt , is solved by integral 3.462 15,
ρb
∞
e − ρx ρ n −1e a ⎡ ⎛ ρb ⎞⎤
∫ (ax + b)
0
n
dx =
a n ⎢Γ⎜ − n + 1, a ⎟⎥ ⇒ by setting
⎣ ⎝ ⎠⎦
e y − x Γ(− 1, y − x )
ρ = 1, b = y − x, a = 1, x = t , n = 2 , and applying Formula
8.352 3, Γ(− n, x ) , which leads to the result.
• Second form hypothesis: solve above relation,
∞ − xt
Ei(± x ) = ± e ± x ∫
e
dt by setting x = x + y in
0
1m t
∞ ∞ ( x + y )t
e − xt x = x + y x + y e −
Ei(+ x ) = e ∫ dt = e ∫ x
dt for
0
1− t 0
1− t
solution Ei( x + y ) ,
and by setting x = x − y in
e − xt x = x − y x − y e − ( x − y )t
∞ ∞
Ei(+ x ) = e ∫ dt = e ∫
x
dt , for
0
1− t 0
1− t
solution Ei( x − y ) .
∞
e −t
o Equivalent form, e x ± y ∫ dt , follows
0
x ± y − t
from same approach by switching
parameters of x ± y in
numerator/denominator,

—Sect. 8.212—
 206

ƒ Verify Ei( x ± y ) solutions by solving the

integral in each,
∞ − ( x ± y )t ∞
x± y e e −t
e ∫ dt = e ∫
x± y
dt , by
0
1− t 0
x± y −t
applying derived integral, Sect. 3.462,
above,
ρb
∞ −
e − ρx n ρ
n −1
e a ⎛ ρb ⎞
∫0 (b − ax )n dx = (− 1) Γ⎜ − n + 1, − ⎟
⎝ a ⎠
n
a
with parameters,
ρ = x ± y (or ρ = 1), a = b = n = 1(or b = x ± y ), x = t and
applying Formula 8.352 3, Γ(− n, x ) .

Ei[− ( x + y )] = Similar to forms in Ei( x + y ) , above, by substituting

⎡ 1
∞
e −t ⎤ − ( x + y ) in Ei( x + y ) .
e −( x + y ) ⎢− +∫ dt ⎥ =
⎣ x + y 0 (y + x + t)
2
⎦
e −(− x − y )t
∞ ∞
e −t
e − x− y ∫ dt = −e − x − y ∫ dt
0
1− t 0
x+ y+t
[ y > 0, x > 0]

—Sect. 8.212—
 207

⎛ 1 ⎞
t
1∞ −
Ei⎜⎜ ⎟= ⎛1⎞ e x
⎝x± y ⎟⎠ Above, in Ei⎜ ⎟ = e x ∫ dt , set
⎝ x⎠ 0
1− t
e −( x ± y )
1 ∞ −1 1 ∞
t
e −t
∫0 1 − t ∫
x± y
dt = e x± y 1 1
e dt x= and x = , and evaluate each integral (first
0
1
−t x+ y x− y
x± y
form) by formula above in Sect. 3.462,
[y > 0, x > 0] −
ρb
∞ − ρx
n ρ ρb ⎞
n −1
e a ⎛
e
∫0 (b − ax )n dx = (− 1) Γ⎜ − n + 1, − ⎟
⎝ a ⎠
n
a
1
with parameters, ρ = , a = b = n = 1, x = t.
x± y
As such,
⎛ 1 ⎞
Ei⎜⎜ ⎟⎟ =
⎝x+ y⎠
e −( x + y )
1 ∞ −1
t

∫0 1 − t
x+ y
e dt =

1
x+ y
⎡ − x +1 y ⎛ 1 ⎞⎤ ⎛ 1 ⎞
e ⎢− e Γ⎜⎜ 0,− ⎟⎟⎥ = −Γ⎜⎜ 0,− ⎟
⎢⎣ ⎝ x+ y ⎠⎥⎦ ⎝ x+ y ⎟⎠
⎛ 1 ⎞
= Ei⎜⎜ ⎟⎟,
⎝x+ y⎠
and
1 ∞ − ( x − y )−1 t
⎛ 1 ⎞ e
Ei⎜⎜ ⎟⎟ = e ∫
x− y
dt =
⎝x− y⎠ 0
1− t
1
⎡ − 1 ⎛ 1 ⎞⎤
e x − y ⎢− e x − y Γ⎜⎜ 0,− ⎟⎟⎥ =
⎣⎢ ⎝ x − y ⎠⎦⎥
⎛ 1 ⎞ ⎛ 1 ⎞
− Γ⎜⎜ 0,− ⎟⎟ = Ei⎜⎜ ⎟⎟.
⎝ x− y⎠ ⎝x− y⎠
1 ∞
e −t
∫
x± y
• Equivalent form, e dt ,
1
0
x± y
−t
follows by switching roles of
1
in numerator/denominator.
x± y

—Sect. 8.212—
 208

⎛ x⎞ ∞
e − xt
Ei⎜⎜ ± ⎟⎟ = In Formula 8.212 5, Ei(± xy ) = ±e ± xy ∫ dt , set y to
⎝ y⎠ 0
ymt
± ∞ ± ∞
x x
e − xt e −t 1
±e ∫ dt = ± e ∫ and solve.
y y
1 x
dt y
0 y
mt 0 y
mt
x∞
± e −t
• Equivalent form, ± e ∫
y
dt , follows
[ y > 0, x > 0]
x
0 y
m t
by switching roles of
x in numerator/denominator.

( )
1
dt
[Re a ] • ( )
Ei a x :
Ei a x = e a ∫0 e x ln a + ln t <0
x
x
o Change Of variable:
s = e x ln a + ln t = a x + ln t ,
( ) = −e ∫ [Re a ]
1
−a x dt
Ei − a x x
>0 dt
e 0
x ln a
− ln t ds = , s 0 = a x + (−∞) = −∞,
t
s1 = a x + 0 = a x ,
es
e =e e
s ax ln t
= te , t =
ax
x
ea
ax
t
I =e ∫ s ds =
ax

−∞
x
a x
es
∫ se ds = Ei(s )
x a
ea ax
−∞ −∞

= Ei a − 0 = Ei a ( ) x
( ) x

• ( )
Ei − a is similar.
x

—Sect. 8.212—
209
 210

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
Change Formula 8.240.1 to read: ∞

By definition of Γ (a , x ) = ∫t
α −1
e − t dt , so that,
x
li( x ) = ∫ = Ei(ln x ) = −Γ(0,− ln x ).
dt x
∞ −t
∞
− Γ (0,− ln x ) = − ( )
ln t e
0

Or, add to Formula 8.241.1: ∫ t

− ln x
dt = − Ei − t
− ln x
=
− Γ(0,− ln x )
− [Ei (− ∞ ) − Ei (lnx )] = Ei (ln x ) = li ( x )
by Formula 2.33.18,
e − βx
n
( )
Ei − βx n
with parameters, [β = n = 1] .
∫ x dx =
n
NOTE: cf. Formula 8.359.2,
⎛ 1⎞
Γ⎜ 0, ln ⎟ = Γ(0,− ln x ) = −li( x ).
⎝ x⎠
¾ Verified math.com

—Sect. 8.240—
 211

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
Formula 8.241.1 Integral,
ln x t
t=∞
I = −Ei(− t )
e
After ∫ dt , =
−∞
t t = − ln x
add formula, − [Ei(− ∞ ) − Ei(lnx )] = li(x )
∞
e −t
− ∫ dt
− ln x
t
( )
Formula 8.241.3, li a x x ln a
et
dt = Ei(t )
x ln a

• Add formula, • ∫
−∞
t −∞
=

( ) ( )
x ln a t
e Ei( x ln a ) − 0 = li e x ln a = li a x
∫−∞ t dt by 2.325.1,
8.211.1 and relation, e x ln a = a x .
• Change present form to: • Change submitted as erratum 3/26/07.
( ) Per e-mail exchange with GR7 Editor,
x t
a
li a = ∫ dt
x
Prof. Alan Jeffrey, 3/27-28/2007, agreed
−∞
t
that original formula,

( )
x
1 at
ln a −∫∞ t
li a =
x
dt , was incorrect;

( ) ∫ at
x t
o correct result is li a x = dt since,
−∞

e t ln a
x x
at
∫t
−∞
dt = ∫ t dt =
−∞

Ei(t ln a ) −x ∞ = Ei( x ln a ) − 0 =
( ) ( )
li e x ln a = li a x
by 8.211.1 and 8.240.1.
NOTE: This erroneous formula has been
listed for over 60 years in G & R (Russian
and English translations), and shows why
listed integrals should be checked by hand or
computer.

—Sect. 8.241—
 212

li( xy ) = Similar to Formula 8.241.1, by substituting

ln xy ∞ xy for x. The integral term,
et e −t
∫ dt = − ∫ dt = ∞

∫e
t t −t
−∞ − ln xy ln tdt =
∞ −ln xy
1
xy ln ln − ∫ e −t ln tdt 1 t=∞
xy −ln xy ln xy ln ln + Ei(− t ) =
xy t = − ln xy
[xy > 1]
1
ln xy ln ln − li(xy )
xy
by integration by parts, with
u = ln t , dv = e − t dt. Thus,
li( xy ) =
1 ⎡ 1 ⎤
xy ln ln − ⎢ xy ln ln − li( xy )⎥ =
xy ⎣ xy ⎦
li( xy ).
¾ Verified math.com
1
• Similar to Formula 8.241.2 by substituting
li( xy ) = xy ∫
dt
; xy for x, and using change of variable on
0
ln xy + ln t
1
each integral,
=
xy
+ xy ∫
dt
; s = ln xy + ln t ;
ln xy 0 [ln xy + ln t ]2
s = ln xy − ln t ,
∞
1 dt each resulting in final form of
= xy ∫
ln xy − ln t t 2 Ei (ln xy ) = li(x).
1

[xy < 1]. ƒ Last integral involves relation,

ln xy
Γ(− 1,− s ) =
−∞
ln xye s ln xy
Ei(s ) −
−∞ s −∞
by new calculations given in Sect. 8.352,
below, for Γ(− n,− x ).
NOTE: integrals with different algebraic
arguments, such as
⎛x⎞
li (x + y ) , li ⎜⎜ ⎟⎟, involves substituting
⎝ y⎠
arguments for x in 8.241.2.
NOTE: see Sect. 4.281, above, for other li(●)
integrals.
¾ Verified math.com

—Sect. 8.241—
213
 214

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
Formula 8.250.4 See Sect. 8.359, below. Formula also given in
1972 printing of
Change
erfc( x ) = 1 − erf (x ) Handbook of Mathematical Functions with Formulas,
to Graphs, and Mathematical Tables. Milton
∞ Abramowitz and Irene A. Stegun, (Eds.). Chapter
erfc( x ) = 1 − erf (x ) =
2
∫e
−t 2
dt 7, “Error Function and Fresnel Integrals,” sect.
π x 7.1.2

—Sect. 8.250—
 215

NOTE: All formulas, in final form, follow from definition and integral representation of
probability integral, 8.250.1 and 8.251.1,
⎛1 ⎞
x x 2 −t γ ⎜ , x2 ⎟
erf ( x) = Φ ( x ) =
2 1 e
dt = ⎝
2 ⎠,
∫ e −t dt = ∫
2

π 0 π 0 t π
by 3.381.8,
m +1
( ) ( )
u
ν −1
∫0 x e dx = nβ γ ν , βu
m − βx n n
ν= [ u > 0, Re ν > 0, Re n > 0, Re β > 0 ]
n

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)

x2
γ (v, β u n ) m +1
e − ty
[Re y ]
2 u

Φ ( xy ) =
y
∫x e =
m − βx dx

π ∫0 t
,ν = > 0 to
n
dt 2
>0 Use integral
0
nβν n
get,
⎛1 ⎞ ⎛1 ⎞
γ ⎜ , y2 x2 ⎟ γ ⎜ , y2 x2 ⎟
I=
y ⎝2 ⎠ = ⎝2 ⎠ = Φ ( xy )
π (y ) π
1
2 2

¾ Verified on math.com

—Sect. 8.252—
 216

Hypothesis: In Formula 8.252.1,

t2 x
⎛ x⎞ x −
Φ( xy ) =
2 y −t 2 y 2 1
Φ⎜⎜ ⎟⎟ =
2
∫e
y2
dt ∫
π 0
e dt , set y = , and use derived
⎝ y⎠ y π 0
y
[ ]
Re y 2 > 0 formula 3.381 8,
( ) γ (ν , βu )
u
−1
∫x e − βx dx = nβ ν
n
m n

m +1
ν= [ u > 0, Re ν > 0, Re n > 0, Re β > 0 ]
n
with parameters ,
⎡ 1 1⎤
⎢m = 0, n = 2, β = y 2 ,ν = 2 ⎥.
⎣ ⎦
Then ,
⎡ 1 ⎛ 1 ⎞2 2 ⎤ ⎡ 1 ⎛ x ⎞2 ⎤
2 γ ⎢ , ⎜⎜ ⎟⎟ x ⎥ γ ⎢ , ⎜⎜ ⎟⎟ ⎥
⎣⎢ ⎝ ⎠ ⎦⎥ = ⎣⎢ ⎝ ⎠ ⎦⎥
x −t 2 y 2 y
∫0 =
y2
e dt 1
1
⎛ 1 ⎞2 2
2⎜⎜ 2 ⎟⎟ y
⎝y ⎠
so that
⎡ ⎛ 1 x2 ⎞ ⎤ ⎛ 1 x2 ⎞
⎢ γ ⎜⎜ , 2 ⎟⎟ ⎥ γ ⎜⎜ , 2 ⎟⎟
⎛x⎞
= ⎝
2 ⎢ ⎝ 2 y ⎠⎥ 2 y ⎠
I= = Φ⎜⎜ ⎟⎟
y π ⎢ 21 ⎥ π ⎝ y⎠
⎢ y ⎥
⎢⎣ ⎥⎦
⎛1 ⎞
by Formula 8.359.4, γ ⎜ , x 2 ⎟ = π Φ( x ) , or from
⎝2 ⎠
⎛1 ⎞
γ ⎜ , x2 ⎟
Φ(x ) = ⎝ ⎠ , Formulas 8.250.1 or 8.251.1.
2
π
¾ Verified on math.com

—Sect. 8.252—
 217

⎛ y ⎞
y
−
t2
⎛ y ⎞
1 Alternative expression for Formula 8.252.6, Φ⎜ ⎟ .
Φ⎜ ⎟ =
⎝ 2x ⎠ x π
∫e
0
4 x2
dt
⎝ 2x ⎠
[Re x ]
x
In Formula 8.252.1, Φ( xy ) =
>0 2 y −t 2 y 2
∫ e dt , set
2

π 0

1
x= y& y= , substitute (including upper limit), and
2x
apply derived formula 3.381.8,
( ) γ (ν , βu )
u
−1
∫x e − βx dx = nβ ν
n
m n

m +1
ν= [ u > 0, Re ν > 0, Re n > 0, Re β > 0 ]
n
with parameters ,
⎡ 1 1⎤
⎢⎣m = 0, n = 2, β = 4 x 2 ,ν = 2 ⎥⎦,

1 1 2 ⎞ γ ⎡⎢ 1 , ⎛⎜ y ⎞⎟ ⎤⎥
2
⎛
t2 γ⎜ , 2 y ⎟
⎠ = ⎢⎣ 2 ⎝ 2 x ⎠ ⎥⎦ ,
y
so that ∫ e dt = ⎝
− 2 2 4x
4x
1
⎛ 1 ⎞
0 ⎛ 1 ⎞2 2⎜ ⎟
2⎜ 2 ⎟ ⎝ 2x ⎠
⎝ 4x ⎠
⎡ ⎛ 1 y2 ⎞ ⎤
⎢γ ⎜ , 2 ⎟ ⎥
1 ⎢ ⎜⎝ 2 4 x ⎟⎠ ⎥
y t2
1 − 2 ⎛ y ⎞
and I =
x π 0 ∫ e dt =
4x

x π ⎢ 1 ⎥
= Φ⎜ ⎟
⎝ 2x ⎠
⎢ x ⎥
⎢⎣ ⎥⎦
⎛1 ⎞
by Formula 8.359.4, γ ⎜ , x 2 ⎟ = π Φ( x ) , or from
⎝2 ⎠
⎛1 ⎞
γ ⎜ , x2 ⎟
Φ(x ) = ⎝
2 ⎠ , Formulas 8.250.1 or 8.251.1.
π
¾ Verified on math.com

⎛ x ± y ⎞ 2 ( x ± y ) −t ⎛x± y⎞
2
x 2⎛ x± y ⎞
⎜ ⎟
Φ⎜ ⎟ solved by experimenting with parameters of
Φ⎜ ⎟=
⎝ a ⎠ ax π 0
∫e ⎝ ax ⎠
dt ⎝ a ⎠
( ) γ (ν , βu )
u
−1
∫x e − βx dx = nβ ν
n

⎡ ⎛ x± y⎞ ⎤ 2 m n

⎢Re⎜ ⎟ > 0⎥ 0

⎢⎣ ⎝ ax ⎠ ⎥⎦ m +1
ν= [ u > 0, Re ν > 0, Re n > 0, Re β > 0 ]
n

Solve integral separately for ± terms.

2
2⎛ x+ y ⎞
⎛ x + y ⎞ −t
x
⎜ ⎟
• The integral in Φ⎜ ⎟, ∫ e ⎝ ax ⎠
dt , is
⎝ a ⎠ 0

—Sect. 8.252—
 218

solved by 3.381.8,
( ) γ (ν , βu )
u
−1
∫x e − βx dx = nβ ν
n
m n

m +1
ν= [ u > 0, Re ν > 0, Re n > 0, Re β > 0 ]
n
with parameters ,
⎡ ⎛x+ y⎞
2
1⎤
⎢m = 0, n = 2, β = ⎜ ⎟ ,ν = ⎥, giving,
⎣⎢ ⎝ ax ⎠ 2 ⎥⎦
⎡ 1 ⎛ x + y ⎞2 2 ⎤
x −t 2 ⎛ x + y ⎞
2 γ ⎢ ,⎜ ⎟ x ⎥
⎜ ⎟ ⎢
⎣ 2 ⎝ ax ⎠ ⎥⎦
∫0 e ⎝ ⎠ dt = ⎡
ax
, so that
2⎤
⎛x+ y⎞ ⎥
2⎢ ⎜ ⎟
⎢ ⎝ ax ⎠ ⎥
⎣ ⎦
⎡1 ⎛ x + y ⎞ ⎤ 2

γ ⎢ ,⎜ ⎟ ⎥
2( x + y ) ⎣⎢ 2 ⎝ a ⎠ ⎦⎥
I= =
ax π ⎡ ⎛ x + y ⎞2 ⎤
2⎢ ⎜ ⎟ ⎥
⎢ ⎝ ax ⎠ ⎥
⎣ ⎦
⎡1 ⎛ x + y ⎞2 ⎤
γ ⎢ ,⎜ ⎟ ⎥
⎢⎣ 2 ⎝ a ⎠ ⎥⎦ ⎛x+ y⎞
= Φ⎜ ⎟
π ⎝ a ⎠
⎛x− y⎞
• Φ⎜ ⎟ is similar.
⎝ a ⎠

Set a = 1 in above,
2( x ± y ) − t
2
x 2⎛ x± y ⎞
⎜ ⎟
Φ(x ± y ) =
x π ∫0
⎝ x ⎠
⎛ x ± y ⎞ 2(x ± y ) −t ⎜⎝ ax ⎟⎠
2
e dt x 2⎛ x± y ⎞

Φ⎜ ⎟=
⎝ a ⎠ ax π 0
∫e dt
⎡ ⎛ x ± y ⎞2 ⎤
⎢Re⎜ ⎟ > 0⎥ ¾ Verified on math.com
⎢⎣ ⎝ x ⎠ ⎥⎦

2⎛ x± y ⎞
2
Set a = 2 in above,
⎛ x ± y ⎞ x ± y −t
x
⎜ ⎟
Φ⎜ ⎟= ∫e ⎝ 2x ⎠
⎛ x ± y ⎞ 2(x ± y ) −t ⎜⎝ ax ⎟⎠
2
dt x 2⎛ x± y ⎞
⎝ 2 ⎠ x π 0 Φ⎜ ⎟=
⎝ a ⎠ ax π 0
∫e dt
⎡ ⎛ x ± y ⎞2 ⎤
⎢Re⎜ ⎟ > 0⎥ ¾ Verified on math.com
⎣⎢ ⎝ 2 x ⎠ ⎥⎦

—Sect. 8.252—
 219

Φ (a + bx ) = Similar to integral above, Φ ( x + y ) ; set

x → bx & y → a
2(a + bx )
bx ⎛ a + bx ⎞ 2
−t 2 ⎜ ⎟

bx π
∫e
0
⎝ bx ⎠
dt ¾ Verified on math.com

⎡ ⎛ a + bx ⎞ 2 ⎤
⎢Re⎜ ⎟ > 0⎥
⎣⎢ ⎝ bx ⎠ ⎥⎦

2(a + bx )
2
a ⎛ a + bx ⎞
−t 2 ⎜ ⎟
=
a π
∫e
0
⎝ a ⎠
dt

⎡ ⎛ a + bx ⎞ 2 ⎤
⎢Re⎜ ⎟ > 0⎥
⎣⎢ ⎝ a ⎠ ⎥⎦

a x −1 •
[Re a ]
First integral:
Φa ( )=
x 2a
∫ e −t
2
a
2
dt 2
>0 a x −1
π 2a 2
∫ e −t dt =
2
0 a

π
e −t (a )
2 x −1 2
x −1 a 0
a
=
π ∫ t
dt
γ ⎢ , (a x −1 ) a 2 ⎥ γ ⎢ , (a x −1 a ) ⎥
⎡1 2 ⎤ ⎡1 2 ⎤

[Re(a ]
0
2a ⎣ 2 ⎦ = ⎣2 ⎦=
)
x −1 2
>0
π 2 (a )
2
π

γ ⎢ , (a x ) ⎥
⎡1 2 ⎤
⎣2 ⎦ = Φ ax ( )
π
by 3.381.8,
( ) γ (ν , βu )
u
−1
∫x e − βx dx = nβ ν
n
m n

m +1
ν= [ u > 0, Re ν > 0, Re n > 0, Re β > 0 ]
n
⎡ 1⎤
with parameters ⎢m = 0, β = a 2 , n = 2,ν = ⎥ .
⎣ 2⎦
• Second integral: in above integral,
x2
e −ty
2

Φ ( xy ) =
y
π ∫0 t
dt ,

Set x = a, y = a x−1 for Φ a x calculation. ( )

¾ verified math.com

—Sect. 8.252—
220
221
 222

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
1− z ∞
⎛1− z ⎞ Apply Formula 3.326.2 with parameters
∫ exp(− ut )t dt [Re u > 0, Re v > 0, Re z < 1]
−z
Γ⎜ ⎟ = vu v v

⎝ v ⎠ ⎡ 1− z ⎤
⎢⎣m = − z; n = v, β = u , γ = v ⎥⎦ to obtain
0

⎛1− z ⎞
Γ⎜ ⎟.
⎝ v ⎠
¾ Verified on Math.com
z −1 ∞
⎛ z −1⎞ Apply Formula 3.326.2 with parameters
∫ exp(− ut ) t dt [Re u > 0, Re v > 0, Re z > 1]
z −2
Γ⎜ ⎟ = vu v v

⎝ v ⎠ ⎡ z − 1⎤
⎢⎣m = z − 2; n = v, β = u, γ = v ⎥⎦ to obtain
0

⎛ z −1⎞
Γ⎜ ⎟.
⎝ v ⎠
¾ Verified on Math.com

—Sect. 8.313—
223
 224

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
Γ( x + a ) = ( x + a − 1)Γ( x + a − 1) ¾ Expansion of Formula 8.331.2
Γ( x + a + 1) • Do integration by parts on
= ∞
x+a Γ( x + a ) = ∫ t x + a −1e −t dt
0

Γ( x + n ) = ( x )n Γ( x ) • nd
2 formula: Do integration by parts on
∞
Γ( x + a + 1) = ∫ t x+ a e −t dt = ( x + a )Γ( x + a )
n n −1
= Γ( x )∏ ( x + n − i ) = Γ( x )∏ ( x + i ),
i =1 i =0 0

where n is a natural number and solve for Γ( x + a )

• 3rd formula, derived from well known
Pochhammer relation,
(a )k = Γ(a + k ) , with a = x.
Γ(a )
¾ Cited, Integrals and Series, Vol. 1,
Elementary Functions, A. P.
Prudnikov, et al., 1986, page 773
and, Mathematica,
http://functions.wolfram.com/GammaBetaErf/Gamma
•4th formula, same as 3rd, in alternative
product notation for reading
Pochhammer forwards or backwards; see
Sect. 8.339, below, for application
Γ(x − a ) = ( x − a − 1)Γ( x − a − 1) ¾ Expansion of Formula 8.331.3
Γ( x − a + 1) • Do integration by parts on
= ∞
x−a Γ( x − a ) = ∫ t x − a −1e −t dt
Γ ( x − n ) = ( x )− n Γ ( x ) =
(− 1) Γ( x )
n
0

(1 − x )n • 2nd formula: Do integration by parts on

∞

Γ( x ) (− 1) Γ(x )
n
Γ( x − a + 1) = ∫ t x − a e −t dt = ( x − a )Γ( x − a )
= n
= n 0

∏ (x − k ) ∏ (k − x )
k =1 k =1
and solve for Γ( x − a )
• 3rd formula, well known, from
(1 − x )n Γ(1 − a )
Γ(n − x ) = Γ(1 − x ) = (1 − x )n−1 Γ(1 − x ),
Pochhammer: (a )k = (− 1)
k
n−x Γ(1 − a − k )
where n is a natural number with x =
Γ( x )
1 − a ⇒ (1 − x )n = (− 1)
n
.
Γ( x − n )
¾ Cited Mathematica,
http://functions.wolfram.com/GammaBetaEr
f/Gamma
• 4th formula, same as 3rd, in alternative
product notation; see Sect. 8.339, below,
—Sect. 8.331—
 225

for application
• 5th formula, derived and applied below
in Sect. 8.339,
⎛ p⎞
under entry Γ⎜⎜ n − ⎟⎟, p < q , from
⎝ q⎠
Pochhammer,
(a )k = Γ(a + k ) , with a = 1 − x, and
Γ(a )
above formula, Γ( x − a ) :

Γ(1 − x + n ) (n − x )Γ(n − x )
(1 − x )n = = ⇒
Γ(1 − x ) Γ(1 − x )
(1 − x )n Γ(1 − x )
Γ(n − x ) =
n−x
= (1 − x )n −1 Γ(1 − x ) if denom. and
last term in (1 − x )n are cancelled.

(n − x )Γ(n − x ) = (− 1)n Γ(x ) , Interrelate above, Γ(n − x ) & Γ( x − n ) , or, by

Γ(1 − x ) Γ( x − n ) Pochhammer,
where n is a natural number (a )k = Γ(a + k ) = (− 1)k Γ(1 − a ) ,
Γ(a ) Γ(1 − a − k )
setting a = 1 − x, k = n :
(1 − x )n = Γ(1 − x + n ) = (− 1)n Γ(x ) , or
Γ(1 − x ) Γ( x − n )
(n − x )Γ(n − x ) = (− 1)n Γ(x ) .
Γ(1 − x ) Γ( x − n )
NOTE: a useful relation which solves many
formulas that follow in this section and Sect.
8.334 such as the following with n = 0.
Γ(n + x )
NOTE: Sect. 8.339, below, relates .
Γ( x − n )
Γ(1 − x )
Γ(− x ) = − [ x ≠ 0,1,2,K] It follows either as a corollary of known
Γ( x + 1)
x formula, Γ(x ) = ,
x
with –x substituted for x, or directly from above
formula Γ( x − a ) ; i.e., evaluate
Γ(0 − x ) where x ≠ 0,1,2,....
NOTE: This well known formula has been
shown to be useful for students evaluating non-
integer arguments of gamma for x < 0. As such,
formula merely summarizes an expression
(“trick”) used in practice to evaluate gamma
functions for x < 0.

—Sect. 8.331—
 226

⎛1⎞
Γ⎜ ⎟
⎛ ⎞
Γ⎜ − ⎟ = − ⎝ ⎠ = −2 π
1 2
EXAMPLE: .
⎝ 2⎠ 1
2
π cosecπ x ¾ One of many possible formulas, presented
Γ(− x ) = −
Γ( x + 1) here, that are based on rearrangement of
reflection formula, Γ( x )Γ(1 − x ) =
π
, GR 8.334.3, and above gamma function
sin π x
relations. Also see Sect. 8.335, Product and
Multiplication Theorems. Proof:
Γ(1 − x )
Γ(− x ) = − by above formula
x
Γ(1 + x )
Γ(x ) = by Formula 8.331.1
x
π
Γ( x )Γ(1 − x ) = , Formula 8.334.3.
sin π x
1⎡ π ⎤ π cos ecπ x
∴ Γ(− x ) = − ⎢ ⎥ =−
x ⎣ Γ( x )sin π x ⎦ Γ( x + 1)
1
since = cos ecπ x.
sin π x
Cited,
http://functions.wolfram.com/GammaBetaErf/Gamma

πx x
π cosec Substitute for x in above formula,
⎛ x⎞ 2 2
Γ⎜ − ⎟ = −
⎝ 2⎠ ⎛ x ⎞ π csc (π x )
Γ ⎜1 + ⎟ Γ(− x ) = − .
⎝ 2⎠ Γ(1 + x )
⎛ πx ⎞
π csc ⎜ ⎟
⎛ x⎞ ⎝ 2⎠
Γ⎜ − ⎟ = −
⎝ 2⎠ ⎛ x⎞
Γ ⎜1 + ⎟
⎝ 2⎠
EXAMPLE:
⎛π ⎞
π csc ⎜ ⎟
⎛ 1⎞
Γ⎜ − ⎟ = − ⎝ 2 ⎠ = − π = −2 π
⎝ 2⎠ ⎛3⎞ 1
π
Γ⎜ ⎟
⎝2⎠ 2

πx ⎛π x ⎞
π cosec π csc ⎜
⎟
⎛ x⎞ 2 ⎛ x⎞ ⎝ 2 ⎠
Γ⎜ ⎟ = In formula, above, Γ⎜ − ⎟ = − ,
⎝2⎠ ⎛ x ⎞ ⎝ 2⎠ ⎛ x⎞
Γ⎜1 − ⎟ Γ ⎜1 + ⎟
⎝ 2⎠ ⎝ 2⎠

—Sect. 8.331—
 227

⎛ x⎞ x ⎛ x⎞ ⎛ x⎞
set Γ⎜1 + ⎟ = Γ⎜ ⎟ , and substitute Γ⎜ ⎟ in
⎝ 2⎠ 2 ⎝2⎠ ⎝2⎠
⎛ x⎞
Γ⎜ − ⎟ .
⎝ 2⎠
⎛π x ⎞ ⎛π x ⎞
π csc ⎜ ⎟ π csc ⎜ ⎟
⎛ x⎞ ⎝ 2 ⎠ ⎝ 2 ⎠
Γ⎜ ⎟ = − =− .
⎝2⎠ ⎛ x⎞ x ⎛ x⎞
Γ ⎜1 + ⎟ Γ⎜ − ⎟
⎝ 2⎠ 2 ⎝ 2⎠
⎛ x⎞ 2 ⎛ x⎞
Then, set Γ⎜ − ⎟ = − Γ⎜1 − ⎟
⎝ 2⎠ x ⎝ 2⎠
by formula Γ(x − a ) above.
⎛π x ⎞
π csc ⎜ ⎟
⎛x⎞ ⎝ 2 ⎠
∴ Γ⎜ ⎟ = − =
⎝2⎠ ⎛ 2 ⎞ ⎡⎛ x ⎞ ⎛ x ⎞⎤
− ⎜ ⎟ ⎢⎜ ⎟Γ⎜1 − ⎟⎥
⎝ x ⎠ ⎣⎝ 2 ⎠ ⎝ 2 ⎠⎦
⎛π x ⎞
π csc ⎜ ⎟
⎝ 2 ⎠
⎛ x⎞
Γ⎜1 − ⎟
⎝ 2⎠

ALTERNATIVE DERIVATION:
In known formula,
π
Γ(x )Γ(1 − x ) = = π csc (π x ) , set
sin (π x )
x x
x= ,1 − x = 1 − , and solve for
2 2
⎛π x ⎞
π csc⎜ ⎟
⎛ x⎞ ⎝ 2 ⎠
Γ⎜ ⎟ = .
⎝2⎠ ⎛ x⎞
Γ⎜1 − ⎟
⎝ 2⎠
EXAMPLE:
⎛3 ⎞
π csc ⎜ π ⎟
⎛3⎞
Γ⎜ ⎟ = ⎝ 2 ⎠ = −π = π .
⎝2⎠ ⎛ 1⎞ −2 π 2
Γ⎜ − ⎟
⎝ 2⎠

πx By formula, Γ( x − a ) , above.
π cosec
⎛ x⎞ 2
Γ ⎜1 − ⎟ =
⎝ 2⎠ ⎛ x⎞ ⎛ x⎞ x ⎛ x⎞
Γ⎜ ⎟ Γ⎜1 − ⎟ = − Γ⎜ − ⎟ .
⎝2⎠ ⎝ 2⎠ 2 ⎝ 2⎠

Since, from above,

—Sect. 8.331—
 228

⎛ πx ⎞ ⎛ πx ⎞
π csc ⎜⎟ π csc ⎜ ⎟
⎛ x⎞ ⎝ 2 ⎠ =− ⎝ 2 ⎠ by
Γ⎜ − ⎟ = −
⎝ 2⎠ ⎛ x⎞ x ⎛ x⎞
Γ⎜1 + ⎟ Γ⎜ ⎟
⎝ 2⎠ 2 ⎝2⎠
πx
π csc
⎛ x⎞ 2 ,
formula Γ⎜ − ⎟ = −
⎝ 2⎠ ⎛ x⎞
Γ ⎜1 + ⎟
⎝ 2⎠

then

⎡ ⎛ πx ⎞ ⎤ ⎛ πx ⎞
⎢ π csc ⎜ ⎟ ⎥ π csc ⎜ ⎟
⎛ x⎞ −x ⎝ 2 ⎠⎥ = ⎝ 2 ⎠.
Γ⎜1 − ⎟ = = ⎢−
⎝ 2⎠ 2 ⎢ x ⎛ x⎞ ⎥ ⎛ x⎞
Γ ⎜ ⎟ Γ ⎜ ⎟
⎢⎣ 2 ⎝ 2 ⎠ ⎥⎦ ⎝2⎠
• ALTERNATIVE DERIVATION:

⎛ x⎞ ⎛ x⎞
Switch the roles of Γ⎜ ⎟ and Γ⎜1 − ⎟ in
⎝2⎠ ⎝ 2⎠
⎛ x⎞
above formula, Γ⎜ ⎟ .
⎝2⎠

⎛π ⎞
π csc ⎜ ⎟
⎛ 1⎞ ⎝ 2⎠ = π.
EXAMPLE: Γ⎜1 − ⎟ =
⎝ 2⎠ ⎛1⎞
Γ⎜ ⎟
⎝2⎠
⎛ x⎞ Obtained from G.S. Carr, Formulas and
2 x −1 π Γ⎜ ⎟ Theorems in Pure Mathematics, 2 ed., 1970,
⎛1− x ⎞ ⎝2⎠
Γ⎜ ⎟= Chelsea Publishing Co., by rearrangement of his
⎝ 2 ⎠ Γ( x ) cos π x formula 2319, page 363, Γ(x ) . Derivation is
2
given there
∞ m −1
x π
from Γ(m )Γ(1 − m) = ∫ dx = , and
0
1+ x sin mπ
21− 2 x π Γ(2 x )
Γ(x ) = .
⎛1 ⎞
Γ⎜ + x ⎟
⎝2 ⎠
1
EXAMPLE: If x = ,
2
⎛1⎞
1
−1
2 2 π Γ⎜ ⎟
⎛1− x ⎞ ⎝ 4 ⎠ = Γ⎛ 1 ⎞.
Γ⎜ ⎟= ⎜ ⎟
⎝ 2 ⎠ Γ⎛ 1 ⎞ cos ⎛ π ⎞ ⎝4⎠
⎜ ⎟ ⎜ ⎟
⎝2⎠ ⎝4⎠

—Sect. 8.331—
 229

NOTE: formula provides another expression

for Γ(x ) , by rearrangement of terms.
⎛ x⎞ ⎛1− x ⎞
21− x π Γ⎜1 − ⎟ In formula above, Γ⎜ ⎟ , set x to
⎛ x −1⎞ ⎝ 2⎠ ⎝ 2 ⎠
Γ⎜ ⎟=
⎝ 2 ⎠ ( x − 1)Γ(1 − x ) cos π x ⎛ x⎞
21− x π Γ⎜1 − ⎟
2 ⎛ x −1⎞ ⎝ 2⎠
2 − x ⇒ Γ⎜ ⎟= ,
⎝ 2 ⎠ ( x − 1)Γ(1 − x ) cos ⎛ π x ⎞
⎜ ⎟
⎝ 2 ⎠
⎛ π x⎞ ⎛π x ⎞
where cos ⎜ π − ⎟ = − cos ⎜ ⎟.
⎝ 2 ⎠ ⎝ 2 ⎠
EXAMPLE: If x =1/2,
1
1− ⎛3⎞
2 2 π Γ⎜ ⎟
⎛ x −1⎞ ⎝ 4 ⎠ = −4Γ⎛ 3 ⎞.
Γ⎜ ⎟= ⎜ ⎟
⎝ 2 ⎠ − 1 cos ⎛ π ⎞ ⎝4⎠
⎜ ⎟
2 ⎝4⎠
⎛ x⎞ ⎛1− x ⎞
2 − x π Γ⎜1 − ⎟ In above formula, Γ⎜ ⎟ , set x to − x
⎛ x + 1⎞ ⎝ 2⎠ ⎝ 2 ⎠
Γ⎜ ⎟=
⎝ 2 ⎠ Γ(1 − x ) cos π x ⎛ x⎞
2 − x −1 π Γ⎜ − ⎟
2 ⎛ x + 1⎞ ⎝ 2⎠ =
⇒ Γ⎜ ⎟=
⎝ 2 ⎠ Γ(− x ) cos ⎛ − π x ⎞
⎜ ⎟
⎝ 2 ⎠
⎛ x⎞
2 − x π Γ⎜1 − ⎟
⎝ 2⎠ ,
⎛π x ⎞
Γ(1 − x ) cos ⎜ ⎟
⎝ 2 ⎠
by substituting
⎛ x⎞ 2 ⎛ x⎞ Γ(1 − x )
Γ⎜ − ⎟ = − Γ⎜1 − ⎟ and Γ(− x ) = − ,
⎝ 2⎠ x ⎝ 2⎠ x
each from above formula, Γ(x − a ) , and where
⎛ π x⎞ ⎛π x ⎞
cos⎜ − ⎟ = cos⎜ ⎟.
⎝ 2 ⎠ ⎝ 2 ⎠
EXAMPLE: If x =1/2,
⎛ 1⎞ ⎛3⎞
1
−
2 2 π Γ⎜1 − ⎟ Γ⎜ ⎟
⎛ x +1⎞ ⎝ 4 ⎠ = ⎝ 4 ⎠ = Γ⎛ 3 ⎞.
Γ⎜ ⎟= ⎜ ⎟
⎝ 2 ⎠ Γ⎛ 1 ⎞ cos ⎛ π ⎞ 2 ⎝4⎠
⎜ ⎟ ⎜ ⎟ 2
⎝2⎠ ⎝4⎠ 2
⎛ x +1⎞
EXAMPLE: If x = 0, Γ⎜ ⎟ = π.
⎝ 2 ⎠
π cosec π x Use Formula 8.334.3,
Γ(x − 1) = .
(x − 1)Γ(1 − x ) Γ(x )Γ(1 − x ) =
π
with x set to 2 − x , 1 − x
sin (π x )

—Sect. 8.331—
 230

set to x − 1 .
π
Γ(2 − x )Γ( x − 1) = =
sinπ (2 − x )
, or
π
− = −π csc (πx )
sin (π x )
π csc (π x )
Γ(x − 1) = − =
Γ(2 − x )
π csc (π x ) π csc (π x )
− = .
(1 − x )Γ(1 − x ) (x − 1)Γ(1 − x )
by relation,
sin ( A − B ) = sinA cosB - cos A sin B,
where A = 2π , B = π x,
and Γ(2 − x ) = (1 − x )Γ(1 − x ) = −( x − 1)Γ(1 − x )
by above formula, Γ(x − a ) .

ALTERNATIVE DERIVATION: Set

Γ( x − 1)Γ(1 − x )
Γ( x − 1) = and use formula
Γ(1 − x )
π
below, Sect. 8.334, Γ( x )Γ(− x ) = −
x sin (π x )
with x set to x − 1, − x set to 1 − x.
Then,
Γ( x − 1)Γ(1 − x )
=
Γ(1 − x )
π π csc (π x )
− =
(x − 1)Γ(1 − x )sin (π x − π ) (x − 1)Γ(1 − x )
since sin (πx − π ) = − sin (πx ) = −
1
.
csc(π x )
• EXAMPLE:
⎛3 ⎞
π csc ⎜ π ⎟
⎛3 ⎞ ⎝2 ⎠ =
Γ⎜ − 1⎟ =
⎝ 2 ⎠ ⎛ 3 − 1⎞ Γ ⎛ 1 − 3 ⎞
⎜ ⎟ ⎜ ⎟
⎝2 ⎠ ⎝ 2⎠
π (− 1)
= π.
⎛1⎞ ⎛ 1⎞
⎜ ⎟Γ⎜ − ⎟
⎝2⎠ ⎝ 2⎠
⎛1 ⎞ 2
2 x −1
π Γ( x ) Exercise
Γ⎜ − x ⎟ = ⎛1 ⎞
⎝2 ⎠ cos πx Γ(2 x ) • Show that Γ⎜ − x ⎟ = −2 π for x =1
⎝2 ⎠
⎛1 ⎞ 4
• Show that Γ⎜ − x ⎟ = π for x =2
⎝2 ⎠ 3

—Sect. 8.331—
 231

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
π Γ(1 − x )
Γ(x )Γ(− x ) = − Γ(− x ) = − by above formula, Sect. 8.331.
x sin π x x
π
Γ( x )Γ(1 − x ) = , Formula 8.334.3
sin π x
⎡ Γ(1 − x ) ⎤
∴ Γ( x )Γ(− x ) = Γ( x )⎢− =
⎣ x ⎥⎦
⎡ π ⎤ π
− Γ(x )⎢ ⎥ =−
⎣ xΓ( x )sin π x ⎦ x sin πx
¾ Cited, Integrals and Series, Vol. 1, Elementary
Functions, A. P. Prudnikov, et al., 1986, page 773
and http://functions.wolfram.com/GammaBetaErf/Gamma

xπ Γ(x + 1) = xΓ( x ) by Formula 8.331.

Γ(1 + x )Γ(1 − x ) =
sin π x π
Γ( x )Γ(1 − x ) = , Formula 8.334.3
sin π x
⎡ π ⎤ xπ
∴ Γ( x + 1)Γ(1 − x ) = xΓ( x )⎢ ⎥ =
⎣ Γ(x )sin π x ⎦ sin π x

Cited, http://functions.wolfram.com/GammaBetaErf/Gamma

π Follows from above formula,

Γ(1 + x )Γ(− x ) = −
sin π x π
Γ( x )Γ(− x ) = − , and Γ( x + 1) = xΓ( x ).
xsin (π x )
Γ( x + 1)Γ(− x ) = xΓ( x )Γ(− x ) =
⎡ π ⎤ π
x ⎢− ⎥ =− .
⎣ xsin (π x ) ⎦ sin (π x )

⎛1 ⎞ ⎛ 1⎞ π
EXAMPLE: Γ⎜ + 1⎟Γ⎜ − ⎟ = − = −π
⎝2 ⎠ ⎝ 2⎠ ⎛π ⎞
sin ⎜ ⎟
⎝2⎠

—Sect. 8.334—
 232

π Γ( x )
Γ(x − 1)Γ(1 − x ) = Γ(x − 1) = by formula Γ(x − a ) above, Sect. 8.331.
(x − 1)sin π x x −1
Γ(1 − x ) = − xΓ(− x ) from Sect. 8.331.
π
Γ( x )Γ(− x ) = − by above formula.
xsin (π x )

∴ Γ( x − 1)Γ(1 − x ) = − Γ( x )Γ(− x ) =
x
x −1
x ⎡ π ⎤ π
− ⎢− ⎥ =
x − 1 ⎣ xsin (π x ) ⎦ ( x − 1)sin (π x )
EXAMPLE:
⎛1 ⎞ ⎛ 1⎞ π
Γ⎜ − 1⎟Γ⎜1 − ⎟ = − = −2π .
⎝2 ⎠ ⎝ 2⎠ ⎛ 1 ⎞ ⎛π ⎞
⎜ − 1⎟ sin ⎜ ⎟
⎝2 ⎠ ⎝ 2⎠
π Γ( x ) π
Γ(− x )Γ( x − 1) = − Γ(− x )Γ( x − 1) = Γ(− x ) =−
x( x − 1)sin π x x −1 x( x − 1)sin (π x )
using above formulas,
π Γ( x )
Γ(x )Γ(− x ) = − , and Γ( x − 1) = .
xsin (π x ) x −1
EXAMPLE:
⎛ 1⎞ ⎛1 ⎞ π
Γ ⎜ − ⎟ Γ ⎜ − 1⎟ = − = 4π .
⎝ 2⎠ ⎝2 ⎠ ⎛ 1 ⎞⎛ 1 ⎞ ⎛ π ⎞
⎜ ⎟⎜ − ⎟sin ⎜ ⎟
⎝ 2 ⎠⎝ 2 ⎠ ⎝ 2 ⎠

—Sect. 8.334—
 233

⎛ 1⎞ ⎛1 ⎞ 2π π
Γ ⎜ x − ⎟ Γ⎜ − x ⎟ = Follows from above formula, Γ( x )Γ(− x ) = − ,
⎝ 2⎠ ⎝2 ⎠ (2 x − 1) cos π x x sin (π x )
1 1
where x is set to x − and − x is set to − x.
2 2

⎛ 1⎞ ⎛1 ⎞ π
Γ⎜ x − ⎟Γ⎜ − x ⎟ = − =
⎝ 2⎠ ⎝2 ⎠ ⎛ 1⎞ ⎛ π⎞
⎜ x − ⎟ sin ⎜ π x − ⎟
⎝ 2⎠ ⎝ 2⎠
− 2π
=
(2 x − 1)[− cos(π x )]
2π
(2 x − 1)[cos(π x )]
where,
⎛ π⎞
sin ⎜ π x − ⎟ = − cos(π x ) by relation,
⎝ 2⎠
sin ( A − B ) = sin A cos B − cos A sin B = 0 − cos A,
π
where A = πx, B = .
2
⎛ 1⎞ ⎛1 ⎞ 2π
NOTE: Γ⎜ x − ⎟Γ⎜ − x ⎟ = may
⎝ 2⎠ ⎝2 ⎠ (2 x − 1)[cos(π x )]
2π sec(π x ) 1
also be expressed as since secA = .
2x − 1 cosA

⎛ 1⎞ π Γ(2 x ) Well known formula, from Gauss Multiplication

Γ( x )Γ⎜ x + ⎟ = Theorem (n = 2 ) , given in Sect. 8.335, below.
⎝ 2⎠ 2 2 x −1
⎛ 1⎞ π Γ(2 x )
• Show that Γ⎜ x − ⎟ =
⎝ 2 ⎠ (2 x − 1)2 2 x − 2
⎛ 1⎞
o Check against x = 1 ⇒ Γ⎜ x − ⎟ = π
⎝ 2⎠
⎛ 1⎞ π
o Check against x = 2 ⇒ Γ⎜ x − ⎟ =
⎝ 2⎠ 2

—Sect. 8.334—
 234

1 ⎞ 2 3− 4 x π ⎡ Γ(2 x ) ⎤ ⎛ 1⎞ ⎛ 1⎞
2
⎛ 1⎞ ⎛ Γ ⎜ x + ⎟ 2Γ ⎜ x + ⎟
Γ⎜ x + ⎟Γ⎜ x − ⎟ =
⎝ 2⎠ ⎝ 2 ⎠ (2 x − 1) ⎢⎣ Γ(x ) ⎥⎦ ⎛ 1⎞
Γ⎜ x − ⎟ = ⎝
2⎠
= ⎝
2⎠
by formula
⎝ 2⎠ 1 2x − 1
x−
2
Γ(x − a ), Sect. 8.331.

⎛ 1⎞
Γ⎜ x + ⎟ is obtained by rearrangement of known
⎝ 2⎠
doubling Formula 8.335.1,
⎛ 1⎞
2 2 x −1 Γ( x )Γ⎜ x + ⎟
⎝ 2⎠
Γ(2 x ) = .
π
2
⎛ 1⎞ ⎛ 1⎞ 2 ⎡ ⎛ 1 ⎞⎤
∴ Γ⎜ x + ⎟Γ⎜ x − ⎟ = ⎢ Γ⎜ x + ⎟ ⎥
⎝ 2⎠ ⎝ 2 ⎠ 2x − 1 ⎣ ⎝ 2 ⎠⎦

2 ⎡ Γ(2 x ) π ⎤ 2 3− 4 x π ⎡ Γ(2 x ) ⎤
2 2

= ⎢ 2 x −1 ⎥ =
2 x − 1 ⎣ 2 Γ( x ) ⎦ 2 x − 1 ⎢⎣ Γ( x ) ⎥⎦

EXAMPLE:
2
⎡ ⎛3⎞
3− 4⎜ ⎟
⎤
⎢ Γ(3) ⎥
3

⎛3 1⎞ ⎛3 1⎞ 2 π
x=
2 ⎝2⎠
Γ⎜ + ⎟Γ⎜ − ⎟ = ⎢ ⎥ = 1.
⎝ 2 2 ⎠ ⎝ 2 2 ⎠ ⎡ ⎛ 3 ⎞ ⎤ ⎢ Γ⎛ 3 ⎞ ⎥
⎢2⎜ 2 ⎟ − 1⎥ ⎢ ⎜⎝ 2 ⎟⎠ ⎥
⎣ ⎝ ⎠ ⎦⎣ ⎦

⎛ x⎞ ⎛x ⎞ 2π π
Γ⎜1 − ⎟Γ⎜ − 1⎟ = In above formula, Γ( x )Γ(− x ) = − , set
⎝ 2 ⎠ ⎝ 2 ⎠ ( x − 2 ) sin π x x sin (π x )
2 x x
x =1− , − x = − 1, and solve
2 2
⎛ x⎞ ⎛ x ⎞ π
Γ ⎜ 1 − ⎟ Γ ⎜ − 1⎟ = − =
⎝ 2⎠ ⎝2 ⎠ ⎛ x ⎞ ⎛π π x ⎞
⎜1 − ⎟ sin ⎜ − ⎟
⎝ 2⎠ ⎝ 2 2 ⎠
2π 2π
− = .
⎛π x ⎞ ⎛π x ⎞
(2 − x )sin ⎜ ⎟ (x − 2)sin ⎜ ⎟
⎝ 2 ⎠ ⎝ 2 ⎠

⎛ 1⎞ ⎛1 ⎞
EXAMPLE: Γ⎜1 − ⎟Γ⎜ − 1⎟ = −2π .
⎝ 2⎠ ⎝2 ⎠

—Sect. 8.334—
 235

(− 1)n π Multiply Γ(n − x )Γ( x − n ) , from Sect. 8.331, above,

Γ(n − x )Γ( x − n ) =
n − x sin π x π
and apply known formula, Γ( x )Γ(1 − x ) = ,
where n is a natural number sin (π x )
Γ(n − x )Γ( x − n ) =
⎡ (− 1)n Γ( x ) ⎤ ⎡ Γ(1 − x )(1 − x )n ⎤ (− 1)n π
⎢ ⎥⎢ ⎥ = .
⎣ (1 − x )n ⎦ ⎣ n − x ⎦ n − x sin (π x )
EXAMPLE: derive known formula Γ(x )Γ(− x ) .
Set n = 0 in

Γ(n − x )Γ( x − n ) ⇒
(− 1)
0
π
=−
π
− x sin (π x ) x sin (π x )
(− 1)n (x )n [Γ(x )]2 From formulas, Sect. 8.331, above.
Γ(n + x )Γ( x − n ) =
(1 − x )n EXAMPLE:
π (− 2 ) π
2
⎛ 1⎞ ⎛1 ⎞ ⎡ ⎛ 1 ⎞⎤
where n is a natural number Γ⎜1 + ⎟Γ⎜ − 1⎟ = = (− 1)⎢Γ⎜ ⎟⎥ = −π .
⎝ 2⎠ ⎝2 ⎠ 2 ⎣ ⎝ 2 ⎠⎦
x[Γ( x )] Γ(1 + x )Γ( x − 1) =
x
[Γ(x )]2 from above formula,
2
Γ(1 + x )Γ( x − 1) =
x −1 x −1
Γ(n + x )Γ(x − n ), with n = 1.
⎛ 3⎞ ⎛3 ⎞ ⎛3 ⎞⎡ ⎤
2
EXAMPLE: Γ⎜1 + ⎟Γ⎜ − 1⎟ = ⎜⎜ 12 ⎟⎟ ⎢Γ⎛⎜ 3 ⎞⎟⎥ = 3 π .
⎝ 2 ⎠ ⎝ 2 ⎠ ⎝ 2 ⎠ ⎣ ⎝ 2 ⎠⎦ 4

Γ(n − x )Γ( x + n ) =
(x )n (1 − x )n π Use Γ(n − x ) from Sect. 8.331, above, known formulas,
n−x sin π x π
Γ(n + x ) = ( x )n Γ( x ) , and Γ( x )Γ(1 − x ) = to
where n is a natural number sin (π x )
obtain,
Γ(n − x )Γ(x + n ) =
(x )n (1 − x )n Γ(x )Γ(1 − x ) = (x )n (1 − x )n
n−x n−x s

EXAMPLE: Derive known formula Γ(1 − x )Γ( x + 1) .

Set n =1 in above formula, Γ(n − x )Γ( x − n ) :
xπ
Γ(1 − x )Γ( x + 1) = .
sin (π x )

—Sect. 8.334—
 236

π This formula extends the known famous reflection

Γ(x )Γ(n − x ) = (1 − x )n π
(n − x )sin π x formula, GR 8.334.3, Γ(x )Γ(1 − x ) = , for
where n is a natural number sin (π x )
0 < x < 1.
Γ(1 − x ) n
Since Γ(n − x ) = ∏ (k − x ) , by Sect. 8.331,
n − x k =1
above, then,
Γ( x )Γ(n − x ) =
⎡∏ (k − x )⎤⎥
n

⎢ k =1
Γ(x )Γ(1 − x )⎢ ⎥=
⎢ n−x ⎥
⎣ ⎦
π π (1 − x )n
∏ (k − x ) =
n
,
(n − x )sin (π x ) k =1 (n − x )sin (π x )
0 < x < 1.
EXAMPLES: Γ( x )Γ(n − x ) can be verified from
Γ( x )Γ(1 − x ) for n =1, and Γ( x )Γ(− x ) for n = 0.
π n =1 π
Γ( x )Γ(n − x ) = ∏ (k − x ) =
n
.
(n − x ) sin (π x ) k = 1 sin (π x )
π n =0 π
Γ( x )Γ(n − x ) = ∏ (k − x ) = −
n

(n − x ) sin (π x ) k = 1 x sin (π x )

NOTE: Γ( x )Γ(n − x ) provides another expression for

π
Γ(n − x ) ⇒ ∏ (k − x ) .
n

(n − x )Γ(x ) sin (π x ) k =1

—Sect. 8.334—
 237

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
NOTE: These important theoretical relations as well as those in Math. Summary form the basis
for many gamma function formulas. See Artin (1964) or Carr (1970) for details.
∞ ⎡
⎛ x⎞ ⎤
2

sin πx = πx∏ ⎢1 − ⎜ ⎟ ⎥
k =1 ⎢⎣ ⎝ k ⎠ ⎥⎦ Important trigonometric relations for derivations of
multiplication & product theorems, and reflection
2 n −1
1 formula. See derivations in Carr, 1970, pages 180 &
= n −1 362.
kπ
∏
n
sin
k =1 n
1 ∞
⎛ k ⎞ −k
x
Alternative equivalent definitions of inverse Γ( x ) .
= xe ∏ ⎜1 + ⎟e
γx

Γ( x ) k =1 ⎝ x⎠ For derivation of reflection formula, Γ( x )Γ(1 − x ),

k 1 1
∞ 1+ GR 8.334.3., begin by solving for
Γ ( x ) Γ (− x )
in
= x∏ x
k =1 ⎛ 1⎞
x
either definition of inverse Γ( x ), and use relation,
⎜1 + ⎟ Γ(1 − x )
⎝ k⎠ Γ(− x ) = .
−x
n −1
⎛ k⎞
1
− nz “Gauss Multiplication Theorem”. From this
∏ Γ⎜ z + ⎟ = Γ(nz )(2π ) n 2
n −1 2

theorem many published results obtain such as the

k =0
⎝ n⎠
doubling formula, Γ(2 x ), Γ(3 x ) and the product
[n = 1,2,K] theorem, Γ(nz ) (GR 8.335), and others. Good
references for gamma function derivations are
Whittaker and G. N. Watson (1934), Artin (1964),
and Carr (1970).
⎛ ⎞ Called “The Product Theorem”.
1 1− n n −1 ⎜ k⎟
nz −
∏ Γ⎜ z + ⎟
Γ(nz ) = n (2π ) k =0 ⎜⎝
2
⎛ k⎞
Solve for Γ(nz ) in above, ∏ Γ⎜ z + ⎟
2
n ⎟⎠ n −1

[n = 1,2,K] k =0
⎝ n⎠
1 1− n n −1 ⎛
⎜ b + k ⎞⎟ Generalizes Γ(nz ) above. This product-sum
Γ(nz + b ) = n
nz + b −
2
(2π )
2
∏ Γ⎜ z +⎜
⎟
k =0
⎝ n ⎟⎠ multiple arguments transformation theorem is cited,
[n = 1,2,K] http://functions.wolfram.com/GammaBetaErf/Gamma,
Formula 06.05.16.0009.01
• If b = 0, Γ(nz + b ) reduces to above Formula,
Γ(nz ) , the “Product Theorem.”
n −1
⎛k⎞ (2π )n −1 See Carr, p 362.
∏ Γ⎜ ⎟ =
k =1 ⎝ n ⎠ n
⎛ k ⎞⎛ k ⎞ (2π )
n −1 n −1 See Carr, p 362.
∏ Γ⎜ ⎟⎜1 − ⎟ =
k =1 ⎝ n ⎠⎝ n⎠ n

—Sect. 8.335—
 238

For n a natural number

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
Formula 8.339.5 Use Γ(n + k − 1 + 1) = Γ( g + 1) = g!= (n + k − 1)! ;
Γ(n + k ) = (n + k − 1)! g = n + k − 1; 2nd formula: do integration by parts on
Γ(n + k + 1) ∞
= Γ(n + k + 1) = ∫ t n+k e −t dt = (n + k )Γ(n + k ),
n+k
0
[n + k ≥ 1, k = 0,1,...]
and solve for Γ(n + k )

Formula 8.339.6
• Use Γ(n − k − 1 + 1) = Γ(g + 1) = g! = (n − k − 1)!;
Γ(n − k ) = (n − k − 1)! g = n − k − 1 ; 2nd formula: do integration by parts
Γ(n − k + 1) on
=
n−k ∞
Γ(n − k + 1) = ∫ t n−k e −t dt = (n − k )Γ(n − k ),
[n − k ≥ 1, k = 0,1,...]
0

and solve for Γ(n − k )

¾ NOTE: Each formula, (n + k − 1)! and

(n − k − 1)!, can be derived by adding the
corresponding (lower and upper)
incomplete gamma functions, derived
below in Section 8.352.

Thus,

Γ(n + k ) = γ (n + k , x ) + Γ(n + k , x ) = (n + k − 1)!

Γ(n − k ) = γ (n − k , x ) + Γ(n − k , x ) = (n − k − 1)!
Γ(n + 1) = n! A classical result. Apply Formula 8.331, Γ(x + 1) , and
Formula 8.339.1, Γ(n ) , for n, a natural number.
⎛ n ⎞ 21− n π (n − 1)! π (n − 2 )!! The first form is well known. The second and simpler
Γ⎜ ⎟= = n −1
⎝2⎠ ⎛ n − 1 ⎞! form is as follows.
⎜ 2 ⎟ 2 2
⎝ ⎠ Derive alternative formula for known formula,
[n = 1,3,5, K] ⎛n⎞
Γ⎜ ⎟, for n a positive odd integer.
⎝2⎠

—Sect. 8.339—
 239

Set n = 2k − 1 for n odd, k = 1,2,K

Then, since n ≥ 1, use
Γ(n − x ) = (1 − x )n −1 Γ(1 − x ) =
n −1
Γ(1 − x )∏ (k − x ), as above, Sect. 8.331 :
k =1

⎛n⎞ ⎛ 1⎞ ⎛ 1 ⎞ k −1 ⎛ 1⎞
Γ⎜ ⎟ = Γ⎜ k − ⎟ = Γ⎜1 − ⎟∏ ⎜ j − ⎟ =
⎝2⎠ ⎝ 2⎠ ⎝ 2 ⎠ j =1 ⎝ 2⎠
n −1
k −1

∏ (2 j − 1) ∏ (2k − 1)
2
π k=
n +1 π
j =1
2

k −1
= k =1
n -1
=
2 22
π (1)(3)(5) K (n − 2) π (n − 2 )!!
n -1
= n -1
.
22 22
⎛ 1⎞
¾ Alternative approach: use Γ⎜ n − ⎟, n ≥ 0 ,
⎝ 2⎠
below, with n = k:
n +1
π (2k − 1)!!
k=
⎛n⎞ ⎛ 1⎞
n = 2 k −1 2
Γ⎜ ⎟ = Γ⎜ k − ⎟ = k −1 =
⎝2⎠ ⎝ 2 ⎠ 2 (2k − 1)
π (n )!! π (n − 2)!!
n −1
= n −1
2 (n )
2
22
since (n )!! = (1)(3)(5) L (n − 2)(n).
⎛ n ⎞ n!! π Exercise. Recall: n!!= 1 ⋅ 3 ⋅ 5L(n − 2 )n , n odd.
Γ⎜ + 1⎟ = n +1 [n = 1,3,5, K]
⎝2 ⎠ 2
2

The first form is well known. The second and simpler

n +1 form is as follows.
π ⎛⎜
n −1⎞ n +1
(− 1)
2
2n ⎟! Derive alternative formula
⎝ 2 ⎠ = (− 2 )
2
⎛ n⎞ π
Γ⎜ −⎟=
(n )!! ⎛ n⎞
⎝ 2⎠ n! for Γ⎜ − ⎟, n a positive odd integer.
[n = 1,3,5, K] ⎝ 2⎠
Set n = 2k − 1 for n odd, k = 1,2, K .
⎛ n⎞ ⎛1 ⎞
Γ⎜ − ⎟ = Γ⎜ − k ⎟. Now, using Formula 8.339.3
⎝ 2⎠ ⎝2 ⎠
with k = n,
⎛1 ⎞ (− 1) π 2
k k
Γ⎜ − k ⎟ = , n ≥ 0.
⎝2 ⎠ ∏ (2 j − 1)
k
j =1

n +1
Substitute back k = and formula results,
2

—Sect. 8.339—
 240

n +1 n +1

⎛ n ⎞ (− 1) 2 π 2 2
Γ⎜ − ⎟ = n +1
=
⎝ 2⎠ 2

∏ (2k − 1)
k =1
n +1 n +1

(− 2) π = (− 2) π .
2 2

(1)(3)(5) K (n ) (n )!!
∏ ( p + qk − q ) ⎛ ⎞ ⎛ p⎞
n

⎛ p⎞ p • Γ⎜ n + ⎟ is derived from Pochhammer-based

Γ⎜ n + ⎟ = k =1 n
Γ⎜⎜ ⎟⎟ ⎝ q⎠
⎝ q⎠ q ⎝q⎠ formula, Sect. 8.331, above,
[n = 0,1,2, K , p < q ] Γ(x + n ) = (x )n Γ( x ), where n is a natural number,
⎛ 1⎞
Γ⎜ n + ⎟ is given in GR as 8.339 2 x→
p
, p < q,
⎝ 2⎠ q
∏ (3k − 2 )
n
so that,
⎛ 1⎞ ⎛1⎞
Γ⎜ n + ⎟ = k =1 n Γ⎜ ⎟ Γ(x + n ) =
⎝ 3⎠ 3 ⎝3⎠ p
x=

2 ⎞ k∏=1(3k − 1) ⎛ 2 ⎞
n

Γ(x )[x( x + 1)( x + 2 )K ( x + n − 1)] =

q
⎛
Γ⎜ n + ⎟= Γ⎜ ⎟
⎝ 3⎠ 3n ⎝3⎠ ⎛ p ⎞ ⎡⎛ p ⎞⎛ p ⎞⎛ p ⎞ ⎛p ⎞⎤
Γ⎜⎜ ⎟⎟ ⎢⎜⎜ ⎟⎟⎜⎜ + 1⎟⎟⎜⎜ + 2 ⎟⎟ L ⎜⎜ + n − 1⎟⎟⎥
( )
n

⎛ 1 ⎞ k∏=1 4k − 3 ⎛ 1 ⎞ ⎝ q ⎠ ⎣⎝ q ⎠⎝ q ⎠⎝ q ⎠ ⎝q ⎠⎦
Γ⎜ n + ⎟ = Γ⎜ ⎟ n
⎝ 4⎠ 4n ⎝4⎠ ( p + qk − q )
n
⎛p ⎞ ⎛ p⎞ ∏ ⎛ p⎞
∏ (4 k − 1) = ∏ ⎜⎜ + k − 1⎟⎟Γ⎜⎜ ⎟⎟ = k =1
n

⎛ 3⎞ ⎛3⎞ n
Γ⎜⎜ ⎟⎟
Γ⎜ n + ⎟ = k =1 n Γ⎜ ⎟ k =1 ⎝ q ⎠ ⎝q⎠ q ⎝q⎠
⎝ 4⎠ 4 ⎝4⎠ p
• Remaining formulas: substitute values for in
q
∏ ( p + qk − q )
n

⎛ p⎞ ⎛ p⎞
Γ⎜ n +⎟ = k =1 n Γ⎜⎜ ⎟⎟ [n = 0,1,2, K , p < q ]
⎝ q⎠ q ⎝q⎠
¾ Cited, Integrals and Series, Vol. 1, Elementary
Functions, A. P. Prudnikov, et al., 1986, page 773
and, Mathematica,
http://functions.wolfram.com/GammaBetaErf/Gamma

—Sect. 8.339—
 241

⎛p ⎞ (− 1)n q n Γ⎛⎜ p ⎞⎟ • Derived from Pochhammer formula, Sect. 8.331,

Γ⎜ − n⎟ = ⎜ ⎟ above,
∏ (qk − p ) ⎝ q ⎠
n
⎝q ⎠
k =1
Γ(x − n ) =
(− 1)n Γ(x ) , where n is a natural number,
[n = 0,1,2, K , p < q ] (1 − x )n
⎛1 ⎞ x→
p
, p < q; note that
Γ⎜ − n ⎟ is given in GR as 8.339 3
⎝2 ⎠ q
⎛ 1 ⎞ (− 1) 3 n ⎛ 1 ⎞
n
(1 − x )n = (1 − x )(2 − x )L (1 − x + n − 1) = ∏ (k − x ).
n
Γ⎜ − n ⎟ = n Γ⎜ ⎟
⎝ 3 ⎠ ∏ (3k − 1) ⎝ 3 ⎠ k =1

(− 1)n Γ⎛⎜⎜ p ⎞⎟⎟

k =1
p
(− 1)n 3n
x=
⎛2 ⎞
Γ⎜ − n⎟ = n
⎛2⎞
Γ⎜ ⎟ Thus, Γ( x − n ) =
q
⎝q⎠ = (− 1)n q n Γ⎛⎜ p ⎞⎟
⎝ 3 ⎠ ∏ (3k − 2 ) ⎝ 3 ⎠ ⎛ p⎞
n ⎜q⎟
( ) ⎝ ⎠
n

k =1 ∏ ⎜⎜ k − ⎟⎟
q⎠
∏ kq − p
k =1 ⎝ k =1
⎛ 1 ⎞ (− 1) 4 n ⎛ 1 ⎞
n
Γ⎜ − n ⎟ = n Γ⎜ ⎟ •
p
⎝ 4 ⎠ ∏ (4k − 1) ⎝ 4 ⎠ Remaining formulas: Substitute values for
q
in
k =1

⎛ 3 ⎞ (− 1) 4 n ⎛ 3 ⎞
n
Γ⎜
⎛p ⎞
− n⎟ =
(− 1)n q n Γ⎛⎜ p ⎞⎟ [n = 0,1,2, K , p < q ]
Γ⎜ − n ⎟ = n Γ⎜ ⎟ ⎜ ⎟
∏ (qk − p ) ⎝ q ⎠
n
⎝ 4 ⎠ ∏ (4k − 3) ⎝ 4 ⎠ ⎝q ⎠
k =1
k =1
¾ Cited,
http://functions.wolfram.com/GammaBetaErf/Gamma

n • New formula follows from formula in Sect. 8.331,

∏ (qk − p )
Γ(n − x ) , with ( p < q ), substituted for x, where
⎛ p⎞
p
⎛ p⎞
Γ⎜ n − ⎟ = k =1 Γ⎜ 1 − ⎟ q
⎝ q ⎠ (nq − p )q n − 1 ⎝ q⎠
n is a natural number.
[ n = 0,1,2,..., p < q] Derivation:
• From Pochhammer relation and formula,
For n = 0 , Γ( x − a ), Sect. 8.331, above:
⎛ p⎞ q ⎛ p⎞
Γ⎜⎜ − ⎟⎟ = − Γ⎜⎜1 − ⎟⎟ (1 − x )n = Γ(1 − x + n ) = (n − x )Γ(n − x ) ⇒
⎝ q⎠ p ⎝ q⎠ Γ(1 − x ) Γ(1 − x )
⎡p ⎤ Γ(n − x ) =
⎢ ∉N⎥ n
(k − x )
(1 − x )n Γ(1 − x ) ∏
⎣q ⎦
= k =1
Γ(1 − x ).
n−x n−x
• Second formula: set n = 0 in above formula and

—Sect. 8.339—
 242

0
remove ∏ (•) term.
k =1

⎛ p⎞ p
EXAMPLE: Derive Γ⎜ n − ⎟, p < q : set x → ,
⎝ q⎠ q
and simplify, to give:
⎛ p⎞ n ⎛ p⎞
Γ⎜ 1 − ⎟ ∏ ⎜k − ⎟
⎛ p⎞ ⎝ q ⎠k = 1 ⎝ q⎠
Γ(n − x ) = Γ⎜ n − ⎟ = =
⎝ q⎠ p
n−
q
n
∏ (qk − p )
k =1 ⎛ p ⎞, n = 0,1,2,... p < q.
⎟Γ⎜ 1 −
(nq − p )q n − 1 ⎝ q ⎠
¾ NOTE: There is a temptation to simplify by
p
canceling the last product term, n − , with the
q
same denominator, but that action losses the
embedded calculation for n = 0 , which is not
desirable so as to conform to Γ(n + x ) & Γ( x − n )
models, above.
p
• Substitute values for in above formula,
⎛ 1⎞ π (2n − 1)!! q
Γ⎜ n − ⎟=
⎝ 2⎠ 2 − 1 (2n − 1)
n n
∏ (qk − p )
∏ (3k − 1) ⎛ p⎞
n
⎛ p⎞ [ n = 0,1,2,..., p < q]
⎛ 1⎞ ⎛2⎞ Γ⎜ n − ⎟ = k =1 Γ⎜ 1 − ⎟
Γ⎜ n − ⎟ = kn=−11 Γ⎜ ⎟ ⎝ q ⎠ (nq − p )q n − 1 ⎝ q⎠
⎝ 3 ⎠ 3 (3n − 1) ⎝ 3 ⎠
∏ (3k − 2 )
n

⎛ 2⎞ ⎛1⎞ ¾ ⎛ 1⎞ π (2n − 1)!! gives another proof

Γ⎜ n − ⎟ = kn=−11 Γ⎜ ⎟ Γ⎜ n − ⎟=
⎝ 3 ⎠ 3 (3n − 2 ) ⎝ 3 ⎠ ⎝ 2 ⎠ 2 n − 1 (2n − 1)

∏ (4k − 1)
n
of Formula 8.338 2,
⎛ 1⎞ ⎛3⎞ ⎛ 1⎞
Γ⎜ n − ⎟ = kn=−11 Γ⎜ ⎟
⎝ 4 ⎠ 4 (4 n − 1) ⎝ 4 ⎠ Γ⎜ − ⎟ = −2 π for n = 0.
⎝ 2⎠
∏ (4k − 3)
n

⎛ 3⎞ ⎛1⎞
Γ⎜ n − ⎟ = kn=−11 Γ⎜ ⎟
⎝ 4 ⎠ 4 (4 n − 3) ⎝ 4 ⎠

∏ ( p − qk ) ⎛ p
n
⎛ p⎞ ⎛ p−q ⎞
⎛ p⎞ ⎞ Γ⎜⎜ ⎟⎟ = Γ⎜⎜ + 1⎟⎟ by algebraic rearrangement,
Γ⎜ ⎟ = k =1 Γ⎜⎜ − n ⎟⎟ ⎝ ⎠
q ⎝ q ⎠
⎝q⎠ qn ⎝q ⎠ ∞

⎡p ⎛ p⎞ ⎤ and using known relation Γ( x + 1) = ∫ t x e −t dt = x Γ( x),

⎜ ⎟
⎢⎣ q ∉ N , n = int ⎜⎝ q ⎟⎠ , p > q⎥
⎦
0

⎛ p⎞ ⎛ p ⎞ ⎛ p ⎞
Γ⎜⎜ ⎟⎟ = ⎜⎜ − 1⎟⎟Γ⎜⎜ − 1⎟⎟, so that repeating the
⎝q⎠ ⎝q ⎠ ⎝q ⎠
factorial process, using above Sect. 8.331 formula,

—Sect. 8.339—
 243

Γ(x − a ) = ( x − a − 1)Γ( x − a − 1) , to produce the

Pochhammer relation, Sect. 8.331 above,
n
Γ( x ) = Γ( x − n )∏ (x − k ), with x =
p
k =1 q

⎛p ⎞ n
Γ⎜⎜ − n ⎟⎟∏ ( p − qk )
⎛ p⎞
Γ⎜⎜ ⎟⎟ = ⎝ ⎠ k =1
q
,
⎝q⎠ qn
⎛ p⎞
where n = int ⎜⎜ ⎟⎟, p > q.
⎝q⎠
NOTE: int (•) is the standard “integer” function which
selects the integer-part of a mixed fractional form of a
⎛3⎞ ⎛ 1⎞
real number, such as int ⎜ ⎟ = int ⎜1 ⎟ = 1. We take
⎝ ⎠
2 ⎝ 2⎠
⎛ p⎞
int ⎜⎜ ⎟⎟ = 0, p < q. For a complete technical definition, see,
⎝q⎠
for example, http://functions.wolfram.com/Notations/.

EXAMPLE:
⎛5 ⎞ 2
Γ ⎜ − 2 ⎟∏ (5 − 2k )
⎛ 5 ⎞ n=2 ⎝ 2 ⎠ k =1 ⎛ 3 ⎞⎛ 1 ⎞ ⎛ 1 ⎞
Γ⎜ ⎟ = = ⎜ ⎟⎜ ⎟Γ⎜ ⎟ =
⎝2⎠ ⎝ 2 ⎠⎝ 2 ⎠ ⎝ 2 ⎠ .
2
2
3
π
4

NOTE: formula is designed to generalize the repeated

application of known formula, Γ( x + 1) = x Γ( x ) ,
p
where x = , p > q, although n = 0 returns
q
⎛ p⎞ ⎛ p⎞
Γ⎜⎜ ⎟⎟ = Γ⎜⎜ ⎟⎟, p < q.
⎝q⎠ ⎝q⎠

Γ(1 − x )
(− 1)n q n Γ⎛⎜⎜ n − p⎞
⎟ Using formula Γ(− x ) = − , above, Sect.
⎛ p⎞ ⎝ q ⎟⎠ x
Γ⎜ −⎟=
⎝ q⎠ ∏
n −1
k =0 ( p − qk ) 8.331, with x =
p
,
⎡p ⎛ p⎞ ⎤ q
⎜ ⎟
⎢⎣ q ∉ N , n = int ⎜⎝ q ⎟⎠ + 1 ⎥⎦ ⎛ p⎞
Γ⎜⎜1 − ⎟⎟
⎛ p⎞ ⎝ q⎠
Γ⎜⎜ − ⎟⎟ = , and repeating the factorial process, wit
⎝ q⎠ p
−
q
Γ(x − a + 1)
h formula Γ(x − a ) = , above, Sect. 8.331,
x−a

—Sect. 8.339—
 244

⎛ p⎞
Γ⎜⎜ − ⎟⎟ =
(− 1)n q n Γ⎛⎜ n − p ⎞⎟,
n −1 ⎜ q ⎟⎠
⎝ q⎠ ( ) ⎝
∏ k =0
p − qk

⎛ p⎞
where n = int ⎜⎜ ⎟⎟ + 1.
⎝q⎠
¾ Alternative approach: In Pochhammer relation,
(a )k = a(a + 1)L (a + n − 1) = Γ(a + k ) ,
Γ(a )
p
set a → − , so that,
q
⎛ p⎞
Γ⎜⎜ n − ⎟⎟ n −1
⎛ p⎞ ⎛ p⎞
⎜⎜ − ⎟⎟ = ⎝
q⎠
= ∏ ⎜⎜ n − ⎟⎟ =
⎝ q ⎠n ⎛ p⎞ k =0 ⎝ q⎠
Γ⎜⎜ − ⎟⎟
⎝ q⎠

(− 1)n ∏ ⎛⎜⎜ p − k ⎞⎟⎟ or,

n −1

k =0 ⎝q ⎠
⎛ p⎞
q n Γ⎜⎜ n − ⎟⎟
⎛ p⎞
Γ⎜⎜ − ⎟⎟ = (− 1) n −1 ⎝
n q⎠
.
⎝ q⎠
∏ ( p − qk )
k =0

⎛1⎞
Γ⎜ ⎟
⎛ 5⎞ 9 ⎛1⎞
n=2
⎝3⎠ .
EXAMPLEs: Γ⎜ − ⎟ = Γ⎜ ⎟ =
⎝ 3 ⎠ 10 ⎝ 3 ⎠ ⎛ − 2 ⎞⎛ − 5 ⎞
⎜ ⎟⎜ ⎟
⎝ 3 ⎠⎝ 3 ⎠
⎛ 1 ⎞ n =1
Γ⎜ − ⎟ = − 2 π .
⎝ 2⎠
NOTE: formula is designed to generalize the repeated
Γ( x + 1)
application of the known relation Γ( x ) = for
x
p
x = < 0. It is a
q
special case of formula
⎛ p⎞ q ⎛ p⎞
Γ⎜⎜ n − ⎟⎟ = − Γ⎜⎜1 − ⎟⎟, p < q, when n = 1.
⎝ q⎠ p ⎝ q⎠

—Sect. 8.339—
 245

⎛ p⎞ n
Γ⎜ n + ⎟= Using formula Γ( x + n ) = Γ( x )∏ ( x + n − i ), above,
⎝ q⎠ i =1

∏ ( p − qk ) ∏ ( p + kq − q ) ⎛ p
m n
p
⎞ Sect. 8.331, with x = ,
k =1 k =1
Γ⎜⎜ − m ⎟⎟ q
qm + n ⎝q ⎠ ⎛ p⎞ ⎛ p⎞ n ⎛ p ⎞
Γ⎜⎜ n + ⎟⎟ = Γ⎜⎜ ⎟⎟∏ ⎜⎜ + n − k ⎟⎟ ,
⎡p ⎛ p⎞ ⎤ ⎝ q⎠ ⎝ q ⎠ k =1 ⎝ q ⎠
⎢ q ∉ N , m = int ⎜ q ⎟, p > q ⎥
⎣ ⎝ ⎠ ⎦ ⎛ p⎞
and substituting formula Γ⎜⎜ ⎟⎟ above for
⎝q⎠
⎛ p⎞ ⎛ p⎞
Γ⎜⎜ ⎟⎟, p > q, with n set to m in Formula Γ⎜⎜ ⎟⎟,
⎝q⎠ ⎝q⎠
⎛p ⎞
Γ⎜⎜ − m ⎟⎟ m
⎛ p⎞
Γ⎜⎜ n + ⎟⎟ = ⎝ m + n ⎠ ∏ ( p − kq )∏ ( p + kq − q ),
n
q
⎝ q⎠ q k =1 k =1

⎛ p⎞
where m = int ⎜⎜ ⎟⎟, p > q.
⎝q⎠

NOTE: If m = 0, then formula reduces to known

⎛ p⎞ m=0
formula above, Γ⎜⎜ n + ⎟⎟, p < q , since ∏ ( p − kq ) = 1 ,
⎝ q⎠ k =1
⎛ p⎞
and int ⎜⎜ ⎟⎟ = 0 , implying p < q.
⎝q⎠

EXAMPLE:
⎛ 4 ⎞ m =1 n
Γ ⎜ − 1 ⎟∏ (4 − 3k )∏ (3k + 1)
⎛ 4 ⎞ m =1 ⎝ 3 ⎠ k =1 k =1
Γ⎜ n + ⎟ =
⎝ 3⎠ 3 n +1
⎛1⎞
Γ⎜ ⎟ n
= ⎝n +1 ⎠ ∏ (3k + 1).
3
3 k =1

⎛ 4 ⎞ (4)(7) ⎛ 1 ⎞
If n = 2, Γ⎜ 2 + ⎟ = Γ⎜ ⎟.
⎝ 3⎠ 33 ⎝3⎠
⎛ 4⎞
As a check, Γ⎜ 2 + ⎟ may be done with Formula
⎝ 3⎠
⎛ p⎞ ⎛ 10 ⎞ 28 ⎛ 1 ⎞
Γ⎜⎜ ⎟⎟ above, Γ⎜ ⎟ = Γ⎜ ⎟.
⎝q⎠ ⎝ 3 ⎠ 27 ⎝ 3 ⎠

—Sect. 8.339—
 246

⎛ p − n⎞ = Use formula from above,

Γ⎜ ⎟
⎝q ⎠ ⎛p ⎞
Γ⎜⎜ − n ⎟⎟ =
(− 1) q n
n
⎛ p⎞
Γ⎜⎜ ⎟⎟ [n = 0,1,2..., p < q ]
(− 1)m + n q n − m ∏ (qk − p ) ⎛ p ⎞ ⎠ ∏ (qk − p ) ⎝ q ⎠
n
⎝q
m
k =1
k =1
Γ⎜⎜ − m ⎟⎟
∏ (qk − p ) ⎛ p⎞
n
k =1 ⎝q ⎠ and substituting formula above, Γ⎜⎜ ⎟⎟, p > q, with
⎡p ⎤ ⎝q⎠
⎛ p⎞
⎢ q ∉ N , m = int ⎜ q ⎟, p > q ⎥ ⎛ p⎞
⎣ ⎝ ⎠ ⎦ n set to m in Formula , Γ⎜⎜ ⎟⎟, p > q, then,
⎝q⎠
⎡ ⎛p ⎞ m ⎤
⎢ ⎜Γ ⎜ − m ⎟⎟∏ ( p − kq ) ⎥
⎛p ⎞
Γ⎜⎜ − n ⎟⎟ = n
(− 1) q ⎢ ⎝ q ⎠ k =1
n n
⎥
⎝q ⎠ ⎢ m
⎥
∏ (kq − p ) ⎢ q
⎥
k =1 ⎣⎢ ⎦⎥
⎛p ⎞ m
(− 1) (q )
n n−m
Γ⎜ − m ⎟⎟∏ ( p − kq )
⎜
⎝q ⎠ k =1 ⎛ reverse sense ⎞
= n
⎜⎜ ⎟⎟
(kq − p ) ⎝ of num. ⎠
∏ k =1

(− 1)n (q )n−m Γ⎛⎜⎜ p − m ⎞⎟⎟∏ (− 1)(kq − p )

m

= ⎝q ⎠ k =1 =
n

∏ (kq − p )
k =1

(− 1)n+ m (q )n−m Γ⎛⎜⎜ p − m ⎞⎟⎟∏ (kq − p )

m

⎝q ⎠ k =1 ,
n

∏ (kq − p )
k =1

⎛ p⎞
where m = int ⎜⎜ ⎟⎟, p > q.
⎝q⎠

NOTE: If m = 0, then formula reduces to known

⎛p ⎞ m=0
formula above, Γ⎜⎜ − n ⎟⎟, p < q , since ∏ (kq − p ) =1,
⎝q ⎠ k =1
⎛ p⎞
and int ⎜⎜ ⎟⎟ = 0 , implying p < q.
⎝q⎠

EXAMPLE:

—Sect. 8.339—
 247

m=2

⎛7 ⎞ m=2 n−2 ⎛ 7
∏ (3k − 7 )
⎞ k =1
Γ⎜ − n ⎟ = (− 1) (3) Γ⎜ − 2 ⎟ n
n+ 2
=
⎝3 ⎠ ⎝3 ⎠ (3k − 7 )
∏
k =1 .

(− 1)n+2 (3) n−2 Γ⎛⎜ 1 ⎞⎟ 4

⎝3⎠
n

∏ (3k − 7)
k =1

NOTE: formula is useful for deriving an alternative

⎛ n⎞
formula for Γ⎜ − ⎟ .
⎝ 2⎠

⎛ p⎞ ⎛ p⎞
Γ⎜ n − ⎟= Use formula above, Γ⎜⎜ n − ⎟⎟, p < q ,
⎝ q⎠
⎝ q⎠
∏ (qk − p )
n
⎛ p⎞ n
⎛ p⎞ Γ⎜⎜1 − ⎟⎟ ∏ (qk − p )
k =1
Γ⎜⎜ m − ⎟⎟ = ⎛ p⎞ ⎝ q ⎠ k =1
Γ⎜ n − ⎟ =
q n − m (nq − p ) ∏ k =1 (qk − p ) ⎝ ,
q (nq − p )
m −1
q⎠ n −1
⎝ q⎠
n −1
∏ (qk − p ) ⎛ p⎞ p ⎛ p⎞
(− 1)1 − m k =1 ⎛ p⎞
Γ⎜⎜ m − ⎟⎟ and set Γ⎜⎜1 − ⎟⎟ = − Γ⎜⎜ − ⎟⎟ from formula
qn − m ∏ k =1
m −1
( p − qk ) ⎝ q⎠ ⎝ q⎠ q ⎝ q⎠
Γ( x − a ) = ( x − a − 1)Γ( x − a − 1), above, Sect. 8.331.
⎡p ⎛ p⎞ ⎤
⎢ q ∉ N , m = int ⎜ q ⎟ + 1, p > q ⎥ ⎛ p⎞
⎣ ⎝ ⎠ ⎦ Then use formula above, Γ⎜⎜ − ⎟⎟, p ≠ q,
⎝ q⎠
with n set to m, defined below, but factor first p in
product and change (0, m − 1) limits to give,
m −1
p∏ ( p − qk ) , which cancels with p in above
k =1

p ⎛ p⎞
transform, − Γ⎜ − ⎟.
q ⎜⎝ q ⎟⎠

—Sect. 8.339—
 248

(− 1)m+1 p(q )m−n Γ⎛⎜ m − p⎞ n

⎟ ∏ (qk − p )
⎛
Γ⎜ n −
p⎞ ⎝ q ⎠ k =1
⎟=
p (nq − p )∏ ( p − qk )
m −1
⎝ q⎠
k =1

(− 1)m+1 Γ⎛⎜ m − p⎞
⎟ ∏ (qk − p )
n

=
⎝ q ⎠ k =1
(nq − p )(q )n−m ∏ (− 1)(qk − p )
m −1

k =1

(− 1)m+1 Γ⎛⎜ m − p⎞ n
⎟ ∏ (qk − p )
=
⎝ q ⎠ k =1
(nq − p )(q ) (− 1)m−1 ∏ (qk − p )
m −1
n−m
k =1

⎛ p⎞ n
Γ⎜ m − ⎟ ∏ (qk − p )
=
⎝ q ⎠ k =1
,
(nq − p )(q )n−m ∏ (qk − p )
m −1

k =1

⎛ p ⎞ + 1, p > q.
where m = int ⎜ ⎟
⎝q⎠
Alternative formula simplifies first one.

EXAMPLE:
∏ (2 k − 5 )
n

⎛ 5⎞ m =3
⎛ 5⎞
Γ⎜ n − ⎟ = Γ⎜ 3 − ⎟ k =1
m −1 =
⎝ 2⎠ ⎝ 2 ⎠ 2 (2n − 5) ∏ (2k − 5)
n − 3
k =1 .
π ∏ (2 k − 5 )
n

k =1
2 n −3 (2n − 5)( −3)( −1)
If
⎛ 5⎞ ⎛3⎞ π (−3)(−1)(1)(3) π
n = 4, Γ⎜ 4 − ⎟ = Γ⎜ ⎟ = 1 = .
⎝ 2⎠ ⎝2⎠ 2 (3)(−3)(−1) 2

NOTE: formula reduces to above formula

⎛ p⎞
Γ⎜⎜ n − ⎟⎟, p < q , if m = 1, implying p < q.
⎝ q⎠

⎛ 1⎞ ⎛1 ⎞ An interesting result for π . Multiply each known

Γ⎜ n + ⎟Γ⎜ − n ⎟ = (− 1) π
n

⎝ 2⎠ ⎝2 ⎠ formula for n integer, GR Formulas 8.339 2 & 8.339

3.

⎛ 1⎞ ⎛1 ⎞
Γ ⎜ n + ⎟Γ ⎜ − n ⎟ =
⎝ 2⎠ ⎝2 ⎠
⎡ ⎤
⎡ π n ⎤ ⎢ (− 1)n π 2 n ⎥
⎢ n ∏ (2k − 1)⎥ ⎢ n ⎥ = (− 1)n π
⎦⎢ ⎥
⎢ ∏ (2k − 1) ⎥
⎣2 k =1

⎣ k =1 ⎦
—Sect. 8.339—
 249

EXAMPLE:
⎛ 1⎞ ⎛1 ⎞ n=2 ⎛ 5 ⎞ ⎛ 3 ⎞
Γ⎜ 2 + ⎟Γ⎜ − 2 ⎟ = Γ⎜ ⎟Γ⎜ − ⎟ = π .
⎝ 2⎠ ⎝2 ⎠ ⎝2⎠ ⎝ 2⎠

This answer follows also from known Formula 8.334,

⎛ 1⎞ ⎛1 ⎞ π x =2
Γ⎜ x + ⎟ Γ⎜ − x ⎟ = = π.
⎝ 2⎠ ⎝2 ⎠ cos(πx )

⎛ 1⎞ ⎛1 ⎞ (− 1) 2π Multiply each formula for n integer—use above

n
Γ ⎜ n − ⎟Γ ⎜ − n ⎟ = ⎛ 1⎞ ⎛1 ⎞
⎝ 2⎠ ⎝2 ⎠ 2n − 1 formula, Γ⎜ n − ⎟ , and known formula Γ⎜ − n ⎟ ,
⎝ 2⎠ ⎝2 ⎠
Formula 8.339 2,
⎛ 1⎞ ⎛1 ⎞
Γ ⎜ n − ⎟Γ ⎜ − n ⎟ =
⎝ 2⎠ ⎝2 ⎠
⎡ ⎤
⎢ n ⎥
⎡ π n ⎤ (− 1) π 2
n
(− 1) 2π
n

⎢ n −1 ∏ (2k − 1)⎥ ⎢ n
⎢ ⎥ =
⎣ 2 (2n − 1) k =1 ⎥ 2n − 1
⎢ ∏ (2k − 1) ⎥
⎦
⎣ k =1 ⎦
NOTE: See formula, Sect. 8.331,
⎛ 1⎞ ⎛1 ⎞ 2π
Γ ⎜ x − ⎟Γ ⎜ − x ⎟ = .
⎝ 2⎠ ⎝2 ⎠ (2 x − 1) cos(π x )
EXAMPLE:

⎛ 1⎞ ⎛1 ⎞ ⎛ 1 ⎞ ⎛ 1 ⎞ n=1 (− 1) 2π
1
Γ⎜1 − ⎟Γ⎜ − 1⎟ = Γ⎜ ⎟Γ⎜ − ⎟ = = −2π .
⎝ 2⎠ ⎝2 ⎠ ⎝2⎠ ⎝ 2⎠ 1

⎛ n ⎞ ⎛ n ⎞ (− 1) 2 2π
n +1
Follows from alternative formulas for each formula
Γ⎜ ⎟ Γ ⎜ − ⎟ = given this Section.
⎝2⎠ ⎝ 2⎠ n
⎛ n⎞ ⎛n⎞
Γ ⎜ − ⎟ Γ⎜ ⎟ =
⎝ 2⎠ ⎝2⎠
⎡ ⎤ ⎡ 1−n n +1
⎤
⎢ ⎥ ⎢
∏ (2 j − 1) ⎥⎥
2

⎢ n +1
⎥⎢ 2 π
(− 1)
2 n +1

⎢ (− 2)
2
π ⎥⎢ j =1 ⎥ 2
2π
⎢ n+1 ⎥⎢ ⎥= .
⎢ 2 ⎥⎢ n ⎥ n
⎢
⎢∏ (2 j − 1) ⎥ ⎢
⎥⎢
⎥
⎥
⎣ j =1 ⎦⎣ ⎦

NOTE: may also be derived from above formula,

Sect. 8.334,
n
x=
π π
Γ( x )Γ(− x ) = −
2
=− .
x sin (π x ) ⎛ n ⎞ ⎛ nπ ⎞
⎜ ⎟ sin ⎜ ⎟
⎝2⎠ ⎝ 2 ⎠
EXAMPLE:
—Sect. 8.339—
 250

⎛ 3 ⎞ ⎛ 3 ⎞ n=3 (− 1) 2π 2π
2
Γ⎜ − ⎟Γ⎜ ⎟ = = .
⎝ 2⎠ ⎝2⎠ 3 3

π n =3 2 π 2
Also follows from − =− = π.
⎛ n ⎞ ⎛ nπ ⎞ 3 ⎛ 3π ⎞ 3
⎜ ⎟ sin ⎜ ⎟ sin ⎜ ⎟
⎝2⎠ ⎝ 2 ⎠ ⎝ 2 ⎠

Γ(n + x )
∏ [(2 x − 1) − (2k − 1) ]
1 n Generalizes Form. 8.334 4.
= 2n
2 2

Γ( x − n ) 2 k =1
• In that formula, set
n 1 ⎛ 1⎞ (2 x − 1) , to
2 2

= ∏ [( x − k )( x + k − 1)] x= p+ , or p = x − 1 / 2, ⎜ x − ⎟ =
k =1
2 ⎝ 2⎠ 4
give:
⎛ 1 ⎞
Γ⎜ p + + n ⎟
⎝ ⎠= 1 [ ]
n

2n ∏
4 p 2 − (2k − 1)
2 2

⎛ 1 ⎞ 2 k =1
Γ⎜ p + − n ⎟
⎝ 2 ⎠
1 n ⎡ ⎛ ⎤
2
1⎞
= 2 n ∏ ⎢4⎜ x − ⎟ − (2k − 1) ⎥ .
2

2 k =1 ⎣⎢ ⎝ 2⎠ ⎦⎥
Γ(n + x )
∴
1 n
Γ( x − n ) 2 k =1
[
= 2 n ∏ (2 x − 1) − (2k − 1) .
2 2
]
o Or, from Sect. 8.331, above, solve by
forming ratio of
Γ( x + n ) Γ( x )
= (x )n Γ( x ) ÷ = ( x )n ( x − 1)n .
Γ( x − n ) (x − 1)n
• Second form: by algebraic simplification of the
difference of two perfect squares.
NOTE: new formula appears to be useful for a variety
of real variables x.

EXAMPLE:
¾ By Formulas 8.339.2 & 8.339.3,
⎛ 1⎞
Γ⎜ n + ⎟
⎝ 2⎠ n (2n − 1)!! (2n − 1)!!
= (− 1)
⎛1 ⎞ 22n
Γ⎜ − n ⎟
⎝2 ⎠
¾ By proposed formula, second form,

—Sect. 8.339—
 251

1
Γ(n + x ) n
x=

= ∏ [( x − k )( x + k − 1)] =
2

Γ( x − n ) k =1
n
⎡⎛ 1 ⎞⎛ 1 ⎞⎤
∏ ⎢⎣⎜⎝ 2 − k ⎟⎠⎜⎝ k − 2 ⎟⎠⎥⎦
k =1
n

∏ [(2k − 1)(2k − 1)]

= (− 1) =
n k =1

2n 2n
(− 1)n (2n − 1)!!2(n2n − 1)!! .
2
.

—Sect. 8.339—
252
 253

FORMULA PROOF
Formula 8.350.3 By definition, a clarification which arises
Γ ( z ,0 ) = Γ ( z ) in practice in the evaluation of some
integrals, and confuses some students
Formula 8.350.4 By definition,
Γ(a, ∞ ) = 0 ∞
Γ(a, ∞ ) = ∫ t a −1e −t dt = 0
∞
Formula 8.350.5 By definition,
γ (a,0) = 0 Re a > 0 0
γ (a,0) = ∫ t a −1e −t dt = 0
0

γ (a, ∞ ) = Γ(a ) Re a > 0 By definition,

∞
γ (a, ∞ ) = ∫ t a −1e −t dt = Γ(a )
0

—Sect. 8.350——
 254

FORMULA PROOF
γ * (− n, x ) = x n [n = 0,1,2,K] • From Gautschi, p. 4., Formula 2.2.
γ (a, x )
• In Formula 8.351.1, γ * (a, x ) = x − a , set
Γ(a )
a → −n, apply 8.356.3, noting the asymptotic
limit, lim Γ(− n ) → ±∞ (see a graph of Γ(x) to
n→∞
see why)
o Thus,
γ (− n )
γ * (− n, x ) = x n =
Γ(− n )
⎡ ⎛ Γ(− n, x ) ⎞⎤
x n ⎢1 − lim ⎜⎜ ⎟⎟⎥ = x n (1 − 0 ) = x n
n →∞ ⎝ Γ(− n ) ⎠
⎣ ⎦

—Sect. 8.351—
 255

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
⎡ ⎛ n + k −1 x m ⎞⎤ • γ (n + k , x ) = Γ(n + k ) − Γ(n + k , x ) by
γ (n + k , x ) = (n + k − 1)! ⎢1 − e − x ⎜⎜ ∑ ⎟⎟⎥ G & R Formula 8.356.3
⎣ ⎝ m =0 m! ⎠⎦
Let g = n + k − 1 → γ (n + k , x ) =
[n + k − 1 ≥ 0, k = 0,1,...] •
γ (n + k − 1 + 1, x )
Then γ (n + k , x ) → γ ( g + 1, x ) =
• ,
Γ(g + 1) − Γ( g + 1, x )
o Noting that Γ(g + 1) = g! and
using G & R Formula 8.352.2
for Γ(1 + n, x ) ;
then γ (n + k , x ) results when
g = n + k − 1 is substituted into
γ ( g + 1, x )
∞
xm
NOTE that the infinite sum, ∑ = ex ,
m =0 m!
Form. 1.211.1

Formula 8.352.6 In above, γ (n + k , x ) , set k = 0.

⎡ ⎛ n −1 x m ⎞⎤
γ (n, x ) = (n − 1)! ⎢1 − e − x ⎜⎜ ∑ ⎟⎟⎥
⎣ ⎝ m =0 m! ⎠⎦
[n = 1,2...]
⎡ ⎛ n − k −1 x m ⎞⎤ Proof similar to proof above for γ (n + k , x )
γ (n − k , x ) = (n − k − 1)! ⎢1 − e − x ⎜⎜ ∑ ⎟⎟⎥
⎣ ⎝ m =0 m! ⎠⎦
[n − k − 1 ≥ 0, k = 0,1,...]

—Sect. 8.352—
 256

n + k −1
xm • Set Γ(n + k , x ) = Γ(n + k − 1 + 1, x )
Γ(n + k , x ) = (n + k − 1)!e − x ∑
m =0 m! • Let g = n + k − 1 → Γ(g + 1, x )
[n + k − 1 ≥ 0, k = 0,1,...] • Apply G & R Formula 8.356.2 Æ
Γ( g + 1, x ) = gΓ( g , x ) + x g e − x
o Substitute into G & R Formula
8.352.2 and simplify
• Alternative derivation: as previously
disclosed, in 2.33.10,
Γ(γ , βx n )
∫ x m − βx n
e dx = − γ
, γ = (m + 1)(n) −1 ;
nβ
then, ∫ x n + k −1e −t dt = −Γ(γ , t ), γ = n + k
γ −1
t m t =∞
= −(γ − 1)!e ∑ −t

m = 0 m! t = x
n + k −1
xm
= (n + k − 1)!e − x
∑
m = 0 m!
¾ Same approach useful for remainder
of derivations.
Formula 8.352.7 In above, Γ(n + k , x ) , set k = 0. Formula cited
n −1
xm
Γ(n, x ) = (n − 1)! e − x ∑
in D. Zwillinger, Handbook of Integration
[n = 1,2,...]
m =0 m! (1992), and
http://mathworld.wolfram.com/IncompleteGammaFuncti
on.html, “Incomplete Gamma Function”
n − k −1
xm Proof similar to above proof for Γ(n + k , x )
Γ(n − k , x ) = (n − k − 1)!e − x ∑
m =0 m!
[n − k − 1 ≥ 0, k = 0,1,...]

Derived from indefinite integral,

e − βx e −t
n
Formula 8.352.8
Γ(− n + k , x ) = ∫ x m dx → − ∫ t n−k +1 dt =
(− 1)n−k ⎡Γ(0, x ) − e − x n−k −1 (− 1)m m! ⎤ − Γ(− z , t ), z = n − k . Substitute into
(n − k )! ⎢⎣ ∑ x m +1 ⎥⎦ (− 1)n ⎡Γ(0, x ) − e− x n −1 (− 1)m m! ⎤
m=0 Γ(− n, x ) = ∑ [n = 1,2,...]
n! ⎢⎣ x m +1 ⎥⎦
[n − k − 1 ≥ 0, k = 0,1,...] m=0

where substitute d is n = z, x = t. Take limits on integral,

∞
e−t
−∫ n − k +1
dt in Γ(− z , t )
x
t
Note that Γ(0, x ) = − Ei(− x ) by Sect. 8.212,
above.
In above, Γ(− n + k , x ) , set k = 1.
Formula 8.352.5
Γ(− n + 1, x ) =
(− 1)n−1 ⎡Γ(0, x ) − e − x n−2 (− 1)m m! ⎤ .
(n − 1)! ⎢⎣ ∑
m =0 x m +1 ⎥⎦
[n = 2,3,...]

—Sect. 8.352—
 257

Γ(0, x ) = −Ei(− x ) ¾ Γ(0, x ) is from definition,

Γ(1, x ) = e −x ∞

∫t e dt = Γ(α , x ) ⇒ ± Γ(0,± x ) =
α −1 − t

Γ(2, x ) = e −x
(x + 1) x

Γ(3, x ) = e −x
(x 2
+ 2x + 2 ) e −t
∞

Γ(4, x ) = e − x (x 3
+ 3x 2 + 6 x + 6 ) ∫ t dt = mEi(m x )
±x
±

¾ For 1 ≤ n ≤ 4 , Formula 8.352.7:

∞

∫t e dt = Γ(n, x ) =
n −1 −t

x
n −1
xm
(n − 1)! e − x ∑ [n = 1,2,...]
m =0 m !

NOTE: useful calculations for certain

derivations of incomplete gamma function.
Γ(1,− x ) = e x ∞

∫t
n −1 −t
e dt =
Γ(2,− x ) = e (1 − x )
x
−x

(
Γ(3,− x ) = e 2 − 2 x + x
x
) 2
Γ(n,− x ) = (n − 1)! e
n −1
(− x )
m
x
∑
Γ(4,− x ) = e (6 − 6 x + 3 x − x )
x 2 3
m =0 m!
n −1
xm
= (n − 1)! e x ∑ (− 1)
[n = 1,2,...]
m

m =0 m!
NOTE: in Formula 8.352.7,
n-1
xm
Γ(n, x ) = (n − 1)!e -x ∑ [n = 1,2,...] ,
m =0 m!

set x → − x.
NOTE: check on calculations:
Γ(n,− x ) + γ (n,− x ) = (n − 1)!
∞
e−x e -t
Γ(− 1, x ) = Ei(-x ) +
x
∫x t n+1 dt = Γ(− n, x ) =
Ei(-x ) − x ⎛ 1 1 ⎞ (− 1)n ⎡
(− 1)m mm +!1 ⎤⎥ =
n −1
Γ(− 2, x ) = − +e ⎜ 2 − ⎟ ⎢ Γ (0, x ) − e −x
∑
2 ⎝ 2x 2x ⎠ n! ⎣ m =0 x ⎦
Ei(-x ) − x ⎛ 1 1 ⎞ (− 1) ⎡− Ei(− x ) − e − x (− 1)m m! ⎤ [n = 1,2,...] ,
n −1
n

Γ(− 3, x ) =
1
+e ⎜ 3 − 2 + ⎟
n! ⎢⎣
∑ x m +1 ⎥⎦
6 ⎝ 3x 6x 6x ⎠ m=0

Γ(− 4, x ) = Formula 8.352.3 (G & R 6), for real x,

for 1 ≤ n ≤ 4 .
Ei(-x ) − x ⎛ 1 1 1 1 ⎞
− +e ⎜ 4 − + − ⎟
⎝ 4x 24 x ⎠
3 2
24 12 x 24 x

—Sect. 8.352—
 258

Γ(0,− x ) = − Ei( x ) ∞
e -t
ex
∫ n+1 dt =Γ(− n,− x ) =
-x t
Γ(− 1,− x ) = Ei( x ) −
x (− 1)n ⎡ n −1
m! ⎤
⎢Γ(0,− x ) − e ∑ (− 1)
x m
⎥
Ei( x ) ⎛ 1 1 ⎞ (− x )m+1 ⎦
Γ(− 2,− x ) = −
n! ⎣
+ ex ⎜ 2 + ⎟ m =0

2 ⎝ 2x 2x ⎠ (− 1)n
⎡ n −1
m! ⎤
Ei( x ) x ⎛ 1 1 ⎞
= ⎢ − Ei ( x ) + e x
∑ m +1 ⎥
[n = 1,2,...]
Γ(− 3,− x ) =
1 n! ⎣ m=0 x ⎦
−e ⎜ 3 + 2 + ⎟
6 ⎝ 3x 6x 6x ⎠ for 1 ≤ n ≤ 4 .
Γ(− 4,− x ) =
NOTE: In formula 8.352.3 (G&R 6), for real x,
Ei( x ) ⎛ 1 1 1 1 ⎞ ∞
− + ex ⎜ 4 + + + ⎟ e -t
24 ⎝ 4x 12 x 3
24 x 2
24 x ⎠ ∫x t n+1 dt = Γ(− n, x ) =
(− 1)n ⎡
( )
n −1
(− 1)m mm+!1 ⎤⎥ [n = 1,2,...],
⎢
n! ⎣
Γ 0, x − e −x
∑
m =0 x ⎦
set x → − x .
γ (1, x ) = 1 − e −x x

∫t e dt = γ (n, x ) =
n −1 −t

γ (2, x ) = 1 − e −x
(x + 1) 0

γ (3, x ) = 2 − e − x (x 2 + 2 x + 2) ⎡ ⎞⎤
⎛ n −1 m
(n − 1)! ⎢1 − e − x ⎜⎜ ∑ x⎟⎟ ⎥ [n = 1,2,...]
γ (4, x ) = 6 − e (x + 3x + 6 x + 6 )
−x 3 2
⎣ ⎝ m =0 m! ⎠ ⎦
γ (− n,± x ) not defined for n > 0. for 1 ≤ n ≤ 4 , Formula 8.352.6
• γ (− n,± x ) from definition, Formula 8.350,
γ (α , x ) [Reα > 0] .

γ (1,− x ) = 1 − e x −x 0
γ (n,− x ) = ∫ t n −1 e −t dt = − ∫ t n −1 e −t dt =
γ (2,− x ) = 1 − e x (1 − x ) 0 −x

γ (3,− x ) = 2 − e x (2 − 2 x + x 2 ) (n − 1)! ⎢1 − e x ⎜⎜ ∑ (− x )
⎡ ⎛ ⎞⎤
n −1 m
⎟⎥ =
m! ⎟⎠⎦⎥
γ (4,− x ) = 6 − e x (6 − 6 x + 3x 2 − x 3 ) ⎣⎢ ⎝ m =0
⎡ ⎛ n −1 x m ⎞⎤
(n − 1)! ⎢1 − e x ⎜⎜ ∑ (− 1)m ⎟⎥ (x < 0),
⎣ ⎝ m =0 m! ⎟⎠⎦
for 1 ≤ n ≤ 4 .
Thus, for example,
−x 0
γ (1,− x ) = ∫ e dt = − ∫ e −t dt =
−t

0 −x

⎡ t=0 ⎤
− ⎢ − e −t ⎥ = 1− e .
x

⎣ t = − x⎦
NOTE: in Formula 8.352.6,
⎡ ⎛ n −1 x m ⎞⎤
γ (n, x ) = (n − 1)! ⎢1 − e − x ⎜⎜ ∑ ⎟⎟⎥
⎣ ⎝ m =0 m! ⎠⎦
set x → − x.
NOTE: adding

—Sect. 8.352—
 259

Γ(n,− x ) + γ (n,− x ) = Γ(n ) = (n − 1)! by 8.356.3,

Γ(a, x ) + γ (a, x ) = Γ(a ).

—Sect. 8.352—
 260

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
∞
• Change of variable:
γ (α , x ) = x ∫ exp(− xe )
α −t
− α t dt s = e −t , ds = −e −t dt = − sdt , s0 = 1, s∞ = 0 , to give:
0

[x > 0, Re α > 0]
∞
( ) (e −t )α dt =
∫e
α − x e−t
I=x
0

e − xs (s )
1 α
γ (a,x ) = x α ∫ t a −1e − xt dt [Re a, x > 0]
0 1
− xα ∫ ds = x α ∫ s α −1e − xs ds
0 1
s 0

γ (α , x )
= xα by Formula 3.381 .8,
xα
u
γ (ν , βu n ) m +1
∫x e
m − βx dx
= ,ν =
n

ν
.
0 nβ n
NOTE: a rearrangement of Formula 3.331 1
with re-expression of parameters to hypothesize
⎛ x⎞
forms below, γ (α , xy ), γ ⎜⎜ α , ⎟⎟, γ (α , x ± y )
⎝ y⎠
• 2nd integral, see Sect. 3.331. Useful for
derivatives in Sect. 8.356
¾ Verified math.com
γ (α , x ± y ) = Similar to above integral,
∞

∫ exp(− xe )
∞
(x ± y ) ∫ exp[− (x ± y )e − α t ]dt
α −t γ (α , x ) = x a −t
− αt dt
0
0

[x > 0, Re y > 0, Reα > 0]

¾ Verified math.com
∞
γ (α , xy ) = ( xy ) ∫ exp(− xye− t − α t )dt
α
Similar to above integral,
∞
0 γ (α , x ) = x a ∫ exp(− xe −t − αt )dt
[x > 0, Re y > 0, Reα > 0] 0

¾ Verified math.com
∞
⎛ x⎞ ⎛ ⎞
γ ⎜⎜ α , ⎟⎟ = (xy −1 ) ∫ exp⎜⎜ − e − t − α t ⎟⎟dt
α x Similar to above integral,
∞
⎝ y⎠ 0 ⎝ y ⎠ γ (α , x ) = x a ∫ exp(− xe −t − αt )dt
0

[x > 0, Re y > 0, Re α > 0]

¾ Verified math.com

—Sect. 8.353—
 261

∞
• In Formula 8.353.5,
Γ(α , x ) = e − x ∫ e −t (t + x )
α −1
dt ∞
Γ(α , xy ) = y e ∫ e (t + x )
α − xy − ty α −1
0 dt , set y = 1 and simplify.
[x > 0, Re α > 1] 0
∞ • 2nd integral, see Sect. 3.331. Useful for derivatives
Γ(a,x ) = x α ∫ t a −1e − xt dt , [Re a,x > 0] in Sect. 8.356
1 ¾ Verified math.com
− ∞ −
x t
⎛ x⎞
Γ⎜⎜ α , ⎟⎟ = y −α e y ∫ e y (t + x ) dt
α −1
Similar to above integral,
⎝ y⎠ 0 ∞

[Re y > 0, x > 0, Re α > 1] Γ(α , x ) = e ∫ e (t + x )

−x −t α −1
dt
0

¾ Verified math.com
∞ Similar to above integral,
Γ(α , x ± y ) = e −( x ± y ) ∫ e −t (t + x ± y )
α −1
dt ∞
Γ(α , x ) = e − x ∫ e −t (t + x )
α −1
0 dt
0

[x > 0, Re α > 1] ¾ Verified math.com

—Sect. 8.353—
 262

FORMULA PROOF OUTLINE

Formula 8.356.7 ∞
Do integration by parts on Γ(a + k , x ) = t ∫
a + k −1 −t
e dt ;
Γ(a + k , x ) x

= (a + k − 1)Γ(a + k − 1, x ) + x a+k −1e − x

2nd formula: do integration by parts on
∞
Γ(a + k + 1, x ) = ∫ t a + k e −t dt =
=
1
a+k
[
Γ(a + k + 1, x ) − x a+ k e − x ] x

(a + k )Γ(a + k , x) + x a+k e − x ,
and solve for Γ( a + k , x)
Formula 8.356.8 In above, Γ(a + k , x ) , set k → − k .

Γ(a − k , x )
= (a − k − 1)Γ(a − k − 1, x ) + x a−k −1e − x

=
1
a−k
[
Γ(a − k + 1, x ) − x a−k e − x ]
Formula 8.356.9 x

γ (a + k , x ) Do integration by parts on γ (a + k , x ) = ∫ t a + k −1e −t dt ;

= (a + k − 1)γ (a + k − 1, x ) − x a+k −1e − x
0
2nd formula: do integration by parts on
1
[
γ (a + k + 1, x ) + x a+k e − x ]
x
=
a+k γ (a + k + 1, x ) = ∫ t a + k e −t dt =
0

−x a+k
+ (a + k )γ (a + k , x),
e −x

and solve for γ (a + k , x)

Formula 8.356.10 In above, γ (a + k , x ) , set k → −k .

γ (a − k , x )
= (a − k − 1)γ (a − k − 1, x ) − x a −k −1e − x

=
1
a−k
[
γ (a − k + 1, x ) + x a−k e − x ]
dΓ(a, x )
∞ See below for derivation of
= Γ(a, x ) ln x + x a ∫ t a −1e − xt ln tdt dΓ(a, x )
∞
da 1 = Γ(a, x ) ln x + x a ∫ t a −1e − xt ln tdt .
da
dγ (a, x )
1 1
= γ (a, x ) ln x + x a ∫ t a −1e − xt ln tdt dγ (a, x )
1
da 0 = γ (a, x ) ln x + x a ∫ t a −1e − xt ln tdt is similar
da 0
using Form 3.331.1.

—Sect. 8.356—
 263

dΓ(a, x ) d ∞ a −1 −t
da ∫x
• = t e dt cannot be solved by Leibnitz’s Rule (Try it).
da
∞
We transform Γ(a, x ) by our formula in Sect.3.331, exp − β e + μx dx =
∫ ( x
) β − μ Γ(μ , β ) , rearrange terms, and
0
then apply Leibnitz’s Rule for the derivative of an integral of constant lower and upper limits of integration.
o In above Sect. 3.331 integral, solve for Γ(μ , β ) , set μ = a, β = x, x = t , to give,

d ⎡ ⎤
∞ ∞
I=x a
∫ exp (− xe ) t
exp (at )dt = Γ (a , x ) . Now we can find
dI
da da ⎣ ∫
= ⎢ x exp (− xe ) exp (at )dt a
⎥ by the
t

0 ⎦ 0
product rule and Leibnitz’s Rule, to give:
dI ⎡ ⎤ d a d ⎡ ⎤
∞ ∞
( )
= ⎢ ∫ exp − xe t exp(at )dt ⎥
da ⎣ 0
x + xa (
⎢ ∫ exp − xe exp(at )dt ⎥ =
t
)
⎦ da da ⎣ 0 ⎦
⎛ Γ(a, x ) ⎞ ⎡ ⎤ ⎡∞ ⎤
∞
x a ln x⎜ ⎟ + x a d
⎢∫ exp − xe t
(
exp (at )
)dt ⎥ = Γ (a , x ) ln x + x a d
⎢ ∫ exp − xe exp(at )dt ⎥.
t
( )
⎝ x ⎠
a
da ⎣ 0 ⎦ da ⎣ 0 ⎦
To solve the remaining integral —call it K—use change of variable:
∞ ∞
e − xs s a
s = e t ; ds = e t dt = sdt ; s o = 1, s ∞ = ∞, ln s = t to give, K = ∫ ds = ∫ s a−1e − xs ds . K can now be
1
s 1
∞ ∞
∂ (a−1)ln s − xs
differentiated by Rule #6 (Math. Summary);
dK
=∫
da 1 ∂a
e ( )
e ds = ∫ s a−1e − xs ln sds.
1

dΓ(a, x )
∞
∴ = Γ(a, x ) ln x + x a ∫ s a−1e − xs ln sds. Set s back to t for standard notation.
da 1

dγ (a, x )
1
¾ For derivation of = γ (a, x ) ln x + x a ∫ t a −1e − xt ln tdt , follow same steps but use Formula
da 0

3.331.1 for γ (a, x ) transform to differentiable two-variable function with constant limits of integration.
dγ (a, x ) dΓ(a, x ) Use Leibnitz’s rules to derive. Check your answers on
• =− http://www.quickmath.com/ or other sources. Solution
dx dx
dΓ( x )
x
exp( xy )dy is found in Bers, Calculus, Vol. 2,
d ⎛ 1 ⎞ d
dx ∫0
• ⎜ ⎟ for
dx ⎜⎝ Γ( x ) ⎟⎠
&
dx
p. 810. Other solutions are possible, using methods in
•
d
erf ( x ) &
d
[Ei(x )] & d [li(x )] the present book.
dx dx dx • More difficult derivatives:
d
x
exp( xy )dy (
d Γ γ , βu n ) (
d γ ν , βu n )
dx ∫0
• &
dγ nβ γ dν Γ(ν )
γ (ν , βu n ) = − γ Γ(γ , βu n )
1 d 1 d NOTE: computer solutions for such derivatives typically
ν
• nβ du nβ du include an advanced special function—Meijer’s G
Function, GR Sect. 9.3. Textbooks usually provide
= u m e − βu , whereν , γ = (m + 1)n −1 dΓ(a, x )
n

solutions similar to ours above, , in terms of

da
integrals.

—Sect. 8.356—
 264

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
⎛1 ⎞
γ ⎜ , x⎟ = π Φ x ( ) NOTE: Alternative derivations from Φ a x , Sect. 8.252. ( )
⎝2 ⎠ • Derivation, π Φ ( x ): By definition, from
⎛1 ⎞
Γ⎜ , x ⎟ = π − π Φ ( x) Formula 8.250 1,
⎛1 ⎞
⎝2 ⎠ γ ⎜ , x⎟
Φ ( x)= 2
∫
x
e −t dt = ⎝
2 2 ⎠⇒
γ ⎢ , (a x ) ⎥ = π Φ (a x )
⎡1 2 ⎤ π 0
π
⎣2 ⎦
⎛1 k ⎞ ⎛ k2 ⎞
⎛1 ⎞
γ ⎜ , x⎟ = π Φ x ( )
γ ⎜ , x ⎟ = π Φ⎜⎜ x ⎟⎟ ⎝2 ⎠
⎝2 ⎠ by Formula 3.381 8,
⎝ ⎠
( ) γ (ν , βu )
u
−1
⎛ k2 ⎞ ∫x e − βx dx = nβ ν
n
m n
⎛1 k ⎞
Γ⎜ , x ⎟ = π − π Φ⎜⎜ x ⎟⎟ 0
⎝2 ⎠ ⎝ ⎠ m +1
ν= [ u > 0, Re ν > 0, Re n > 0, Re β > 0 ]
⎛1 k ⎞
⎛ 2⎞
k
⎛ k2 ⎞ n
γ ⎜ ,− x ⎟ = π Φ⎜ ix ⎟ = i π erfi⎜⎜ x ⎟⎟
⎜ ⎟
⎡ 1⎤
⎝2 ⎠ ⎝ ⎠ ⎝ ⎠ with parameters ⎢m = 0, n = 2, β = 1,ν = .
⎣ 2 ⎥⎦
⎛1 ⎞ ⎛ ⎞
( )
k
Γ⎜ ,− x k ⎟ = π − π Φ⎜⎜ ix 2 ⎟⎟ • To establish, π − π Φ x —by Formula
⎝2 ⎠ ⎝ ⎠ 8.356 3, Γ(α ) = γ (α , x ) + Γ(α , x ) , so that, with
⎛ k2 ⎞ 1 ⎛1⎞ ⎛1 ⎞ ⎛1 ⎞
= π − i π erfi⎜⎜ x ⎟⎟, α = , Γ⎜ ⎟ = γ ⎜ , x ⎟ + Γ⎜ , x ⎟ . Solve for
⎝ ⎠ 2 ⎝2⎠ ⎝2 ⎠ ⎝2 ⎠
where, ⎛1 ⎞ ⎛1 ⎞
Γ⎜ , x ⎟ and substitute γ ⎜ , x ⎟ , above, noting
⎛1 ⎞
γ ⎜ ,− x 2 ⎟ = π Φ(ix ) = i π erfi( x) ⎝2 ⎠ ⎝2 ⎠
⎝2 ⎠ ⎛1⎞
[see Math. Summary] Γ⎜ ⎟ = π .
⎝2⎠
NOTE: See —A.P. Prudnikov, et al., Integrals and
Series, Vol. 2: Special Functions, 1986, page 726, “The
Incomplete Gamma Function”:
⎧γ (1 / 2, z )
= π
⎧⎪erf z ⎫⎪ ( )
⎨ ⎨ ⎬
⎩Γ(1 / 2, z ) ⎪⎩erfc z ⎪⎭ ( )
• Last four results: from 3rd result; set:

(a )
k
x 2
→ x ⇒ a = x , or
k x 2

(a )
k
x 2
→ − x ⇒ a = ix , k x 2

and Formula 8.356.3

—Sect. 8.359—
 265

⎡ e−x ⎛ 1 ⎞⎤ 1
⎛ 1 ⎞
Γ⎜ − , x ⎟ = 2 ⎢ − Γ⎜ , x ⎟ ⎥ • From Formula 8.356 2, Γ(α + 1, x ), with α = − .
⎝ 2 ⎠ ⎣ x ⎝ 2 ⎠⎦ 2
Solve for Γ(α , x )
⎛ n ⎞
• See Γ⎜ − , x ⎟ , below, for general case, and
⎝ 2 ⎠
substitute n = 1.
⎡ e−x ⎞⎤ ⎛ 1 ⎞
2
⎛ 1 ⎞ ⎛1 Similar to Γ⎜ − , x ⎟ with x → x 2
Γ⎜ − , x 2 ⎟ = 2 ⎢ − Γ⎜ , x 2 ⎟ ⎥
⎝ 2 ⎠ ⎢⎣ x ⎝2 ⎠⎥⎦ ⎝ 2 ⎠
⎛1 ⎞
Γ⎜ , − x ⎟ = π − πΦ i x ( ) ⎛1 ⎞
In above formula, Γ⎜ , x ⎟ , set x → − x
⎝2 ⎠ ⎝2 ⎠
⎛1 ⎞
Γ⎜ , − x 2 ⎟ = π − π Φ(ix )
⎛1 ⎞
Similar to Γ⎜ ,− x ⎟ = π − π Φ i x ( )
⎝2 ⎠ ⎝2 ⎠
⎛1 ⎞
γ ⎜ ,− x ⎟ = ( )
πΦ i x
⎛1 ⎞
In above formula, γ ⎜ , x ⎟, set x → − x
⎝2 ⎠ ⎝2 ⎠
⎛1 ⎞
γ ⎜ ,− x 2 ⎟ = π Φ(ix )
⎛1 ⎞
Similar to γ ⎜ ,− x ⎟ = π Φ i x ( )
⎝2 ⎠ ⎝2 ⎠
⎛ 1 ⎞ ⎡ ex ⎛1 ⎞⎤ ⎛ n ⎞
Γ⎜ − , − x ⎟ = 2 ⎢ − Γ⎜ , − x ⎟ ⎥ Below, in general case, Γ⎜ − ,− x ⎟ , set n → 1.
⎝ 2 ⎠ ⎣i x ⎝2 ⎠⎦ ⎝ 2 ⎠
⎡ex ⎞⎤ ⎡ ex ⎞⎤
2
⎛ 1 ⎞ ⎛1 ⎛ 1 ⎞ ⎛1
Γ⎜ − , − x 2 ⎟ = 2 ⎢ − Γ⎜ , − x 2 ⎟ ⎥ Similar to above, Γ⎜ − ,− x ⎟ = 2 ⎢ − Γ⎜ , − x ⎟ ⎥
⎝ 2 ⎠ ⎣⎢ ix ⎝2 ⎠⎦⎥ ⎝ 2 ⎠ ⎣i x ⎝2 ⎠⎦
1 ⎡ ⎛1 2⎞ ⎛ 1 2 ⎞⎤ ⎛1 ⎞
⎢γ ⎜ 2 , x ⎟ + Γ⎜ 2 , x ⎟⎥ = x γ ⎜ , x2 ⎟
π ⎣ ⎝ ⎠ ⎝ ⎠⎦ 2 2 ⎝2 ⎠=
∫e
−t 2
dt =
erf ( x ) + erfc(x ) = 1 • π 0 π 2
1 ⎛1 ⎞
γ ⎜ , x 2 ⎟ = erf ( x )
π ⎝2 ⎠
∞
2
∫ dt =
−t 2
e
π x
• ⎛1 ⎞
Γ⎜ , x 2 ⎟
2 ⎝2 ⎠ = 1 Γ⎛ 1 , x 2 ⎞ = erfc( x )
⎜ ⎟
π 2 π ⎝2 ⎠
x ∞ ∞
2 2 2
∫ e dt + ∫ e dt = π ∫
2 2
−t −t −t 2
• e dt = 1 ,
π 0 π x 0
by Formula 3.326 2 and Formula 8.356 3,
Γ(a ) = γ (a, x ) + Γ(a, x ).

¾ Proof that erf (− ∞ ) = Φ (− ∞ ) = −1

By definition,
−∞
2
erf (− ∞ ) = Φ(− ∞ ) = ∫e .
2
−t

π 0

—Sect. 8.359—
 266

0
2
Change limits to read, erf (− ∞ ) = Φ(− ∞ ) = − ∫e
−t 2

π −∞
x
2
Note that the integrand of erf (x ) = ∫e
−t 2
is an even-
π 0
function (see Bers, vol. 1, p. 261)—see graph below—
so that,
−∞ 0 ∞
2 2 2
∫ e dt = − ∫ e dt = − ∫e
2 2
−t −t −t 2
erf (−∞) = dt = −1
π 0 π −∞ π 0

(even-function property on ± ∞ ▼)

NOTE: the erf integral is odd function since

erf (− x ) = −erf ( x ) . Derive this relation.

—Sect. 8.359—
 267

⎛ n ⎞ ⎛ n ⎞ ⎛n ⎞
n
• First integral from Γ(a + k , x ), Sect. 8.356,
[n = 1,3,5,K]
−1
Γ⎜ , x ⎟ = ⎜ − 1⎟ Γ⎜ − 1, x ⎟ + x 2 e − x
⎝ 2 ⎠ ⎝ 2 ⎠ ⎝2 ⎠ n
above, with a → − 1, k = 1
2
⎡ −x ⎤ o See expansion expression, below
⎛ n ⎞ 2 ⎢e ⎛ n ⎞⎥
Γ⎜ − , x ⎟ = − Γ⎜ 1 − , x ⎟ [n = 1,3,5,....] • Second is similar to Γ⎜ −
⎛ 1 ⎞
, x ⎟, above
⎝ 2 ⎠ n⎢ 2 ⎝ 2 ⎠⎥
n

⎣x ⎦ ⎝ 2 ⎠
with x → − x
⎛n ⎞
• Exercise—find γ ⎜ , x ⎟
⎝2 ⎠
⎡ ⎤ • First integral: In 8.356.2, set
⎛ n ⎞ 2⎢ e ⎛ n ⎞⎥
x
Γ⎜ − ,− x ⎟ = − Γ⎜1 − , − x ⎟ n
a → − , x → − x , noting that
⎝ 2 ⎠ n⎢ ⎝ 2 ⎠⎥
n n
2
⎢⎣ (− 1) x
2 2
⎥⎦ n 1 n −1

[n = 1,3,5,....] (− 1)
2
= (− 1) (− 1)
2 2
= ±i for odd integer n.
⎛n ⎞
n
⎛n ⎞ ⎛ n ⎞ ⎛n ⎞
n
−1
Γ⎜ ,− x ⎟ = ⎜ − 1⎟ Γ⎜ − 1,− x ⎟ − (− 1) x • Second is similar to Γ⎜ , x ⎟ , above,
2
2
ex
⎝2 ⎠ ⎝ 2 ⎠ ⎝2 ⎠ ⎝2 ⎠
[n = 1,3,5,....] with x → − x
⎛n ⎞ ⎛n ⎞ ⎛1 ⎞
Γ⎜ , x ⎟ = • Γ⎜ , x ⎟ is a generalization of Γ⎜ , x ⎟ for
⎝2 ⎠ ⎝2 ⎠ ⎝2 ⎠
odd positive integer n.
(n − 2)!! ⎡⎢ 1 Γ⎛ 1 , x ⎞ + e − x x 12 (2 x )s ⎤
n −3
2

⎜ ⎟ ∑ ⎥
s = 0 (2 s + 1)(2 s − 1)!!⎥
• Derivation: In Form. 8.356.5,
Γ(α + n, x ) Γ(α , x ) − x n−1
n −3
⎢⎣ 2 ⎝ 2 ⎠ ⎦ xα + s
2 2
= +e ∑ ,
[n = 1,3,5,...] Γ(α + n ) Γ(α ) s =0 Γ (α + s + 1)

o Solve for Γ(α + n, x )

1
⎛3 ⎞ 1 ⎛1 ⎞
1
o Set α =
Γ⎜ , x ⎟ = Γ⎜ , x ⎟ + e − x x 2 2
⎝2 ⎠ 2 ⎝2 ⎠ n −1
1 o Set n = for
⎛5 ⎞ 3 ⎛1 ⎞ ⎛
⎜3
⎞
Γ⎜ , x ⎟ = Γ⎜ , x ⎟ + e − x x 2 ⎜ + x ⎟ 2
⎟

⎝ 2 ⎠ 4 ⎝ 2 ⎠ ⎜2
⎝
⎟
⎠ ⎛n ⎞ ⎛ 1 n −1 ⎞
1 Γ⎜ , x ⎟ = Γ ⎜ + , x⎟
⎛ 7 ⎞ 15 ⎛ 1 ⎞ ⎛
⎜ 15 5 ⎞
⎝2 ⎠ ⎝2 2 ⎠
Γ⎜ , x ⎟ + e − x x 2 ⎜ + x + x 2 ⎟
⎟
Γ⎜ , x ⎟ =
⎝2 ⎠ 8 ⎝2 ⎠ ⎝ 4
⎜ 2 ⎟
⎠ o Substitute
⎛n⎞ π (n − 2 )!!
Γ⎜ ⎟ = n −1
[n = 1,3,5, K]
⎝ ⎠
2
2 2
derived in Sect. 8.339, above
⎛1⎞
Γ⎜ ⎟ = π from 8.338.2
o
⎝2⎠
⎛n ⎞
• Expressing Γ(α + n, x ) = Γ⎜ , x ⎟ and
⎝2 ⎠
1
factoring out x 2 ,

—Sect. 8.359—
 268

⎛n ⎞
Γ⎜ , x ⎟ =
⎝2 ⎠
⎡ ⎛1 ⎞ ⎤
⎢ Γ⎜ , x ⎟ n −3
1 2 ⎥
⎛n⎞ 2 ⎠
Γ⎜ ⎟ ⎢ ⎝
s
x
+ e −x x 2 ∑ ⎥
⎝ 2 ⎠ ⎢ Γ⎛ 1 ⎞ s =0 ⎛3 ⎞⎥
Γ⎜ + s ⎟ ⎥
⎢ ⎜⎝ 2 ⎟⎠ ⎝2 ⎠⎦
⎣

⎡ ⎛1 ⎞ ⎤
⎢ Γ⎜ 2 , x ⎟ π ⎥
⎢ ⎝ ⎠ + ⎥
⎢
(n − 2)!! ⎢ π ⎥
= n −1 n −3 ⎥
⎢ 1 2
x s ⎥
2 2 ⎢ π e −x x 2
⎢
∑
s =0 ⎛3 ⎞⎥
⎥
Γ⎜ + s ⎟
⎢⎣ ⎝2 ⎠ ⎥⎦
⎡ n −3 ⎤
⎢
(n − 2)!! ⎢Γ⎛ 1 , x ⎞ + π e − x x 21 2
x s ⎥
= n −1
⎢
⎜
⎝ 2
⎟
⎠
∑ = ⎛ 3 ⎞
⎥
⎥
2 2 ⎢
s 0
Γ⎜ + s ⎟ ⎥
⎣ ⎝2 ⎠⎦

Factor
⎛3 ⎞
Γ⎜ + s ⎟ =
⎝2 ⎠
⎛ 1⎞ ⎛ 1 ⎞ ⎛ 2s + 1 ⎞ ⎛ 1⎞
⎜ s + ⎟ Γ⎜ s + ⎟ = ⎜ ⎟ Γ⎜ s + ⎟
⎝ 2⎠ ⎝ 2⎠ ⎝ 2 ⎠ ⎝ 2⎠
and use Formula 8.339 2,
⎛ 1⎞ π (2n − 1)!!
Γ⎜ n + ⎟ = , with n → s,
⎝ 2⎠ 2n
rearranging and simplifying,
⎛n ⎞
Γ⎜ , x ⎟ =
⎝2 ⎠
⎡ ⎛1 ⎞ ⎤
⎢ Γ⎜ 2 , x ⎟ + ⎥
(n − 2)!! ⎢ ⎝ ⎠ ⎥
n −1 ⎢ −x 2
1 n −3
⎥
⎢ 2 ⎥
2 s s
e x x
2 2∑
2
⎢ π π s =0 (2 s + 1)(2s − 1)!!⎥⎦
⎣

=
(n − 2)!! ⎡Γ⎛ 1 , x ⎞⎤ +
n −1 ⎢ ⎜ 2 ⎟⎥
⎣ ⎝ ⎠⎦
2 2
n −3
(n − 2)!! e − x x 1 ⎛⎜ s
2 2 x ⎞⎟⎠
n −3
2
∑ (2s + 1)(2s − 1)!!
s =0
⎝

2 2

—Sect. 8.359—
 269

1
and factoring n −3
,
2 2

⎛n ⎞
Γ⎜ , x ⎟ =
⎝2 ⎠
⎡1 ⎛ 1 ⎞ ⎤
⎢ 2 Γ⎜ 2 , x ⎟ + ⎥
(n − 2)!! ⎢ ⎝ ⎠ ⎥
n −3 ⎢ n −3
⎥
⎢
1 2
(2 x ) s
⎥
⎢e x ∑ (2s + 1)(2 s − 1)!!⎥
2 2 −x 2

⎣ s =0 ⎦
which is final proposed form.
⎛1 ⎞
NOTE: if n = 1, Γ⎜ , x ⎟ = π − π Φ ( x ) by
⎝2 ⎠
−1
above since (−1)!!=1 and ∑ (•) = 0 , so formula
0
is general.
• Examples:
1
⎛3 ⎞ 1 ⎛1 ⎞ −x 2
Γ⎜ , x ⎟ = Γ⎜ , x ⎟ + e x
⎝2 ⎠ 2 ⎝2 ⎠
1
⎛5 ⎞ 3 ⎛1 ⎞ ⎛
⎜3
⎞
Γ⎜ , x ⎟ = Γ⎜ , x ⎟ + e − x x 2 ⎜⎜ + x ⎟⎟
⎟

⎝2 ⎠ 4 ⎝2 ⎠ ⎝2 ⎠
1
⎛ 7 ⎞ 15 ⎛ 1 ⎞ ⎛
⎜ 15 5 ⎞
Γ⎜ , x ⎟ = Γ⎜ , x ⎟ + e − x x 2 ⎜⎜ + x + x 2 ⎟⎟
⎟

⎝2 ⎠ 8 ⎝2 ⎠ ⎝ 4 2 ⎠

⎛n ⎞ ⎡ Γ(α , x ) − x α n −1 xs ⎤ • From Formula 8.356.5, above.

Γ⎜ , x ⎟ = Γ(α + n )⎢ +e x ∑ ⎥
⎣ Γ(α ) s = 0 Γ(α + s + 1) ⎦
⎝k ⎠ ⎛7 ⎞
EXAMPLE: Evaluate Γ⎜ , x ⎟
⎝3 ⎠
⎛n⎞ n ⎛n⎞
n = int ⎜ ⎟, α = − int ⎜ ⎟ ⎛n⎞ 7 1
⎝k⎠ k ⎝k⎠ Set n = int ⎜ ⎟ = 2;α = − 2 = to give,
⎝k⎠ 3 3
⎡n ⎤ 1
⎢⎣ k ≠ 1,2,3K⎥⎦ ⎛4 ⎞
⎛7 ⎞ ⎛ 1 ⎞ 4 ⎛1 ⎞
Γ⎜ , x ⎟ = Γ ⎜ 2 + , x ⎟ = Γ⎜ , x ⎟ + x 3 e − x ⎜ + x ⎟
⎝3 ⎠ ⎝ 3 ⎠ 9 ⎝3 ⎠ ⎝3 ⎠
¾ Simplify Γ(α + n ) , Γ(α + s + 1) by
⎛ p⎞
methods of Section 8.339, Γ⎜⎜ ⎟⎟
⎝q⎠
• See above, Sect. 8.356, Form. 8.356.7.

—Sect. 8.359—
 270

⎛n ⎞ ⎛n ⎞
γ ⎜ , x⎟ = • Complement of above, Γ⎜ , x ⎟ , for odd
⎝2 ⎠ ⎝2 ⎠
⎡ n −3
⎤ positive integer n.
(n − 2)!! ⎢ 1 π Φ( )
1 2
(2 x )s ⎥ • From Formula 8.356 3,
n −3 ⎢ 2
x − e −x x 2
∑
s =0 (2 s + 1)(2 s − 1)!! ⎥ Γ(a ) = γ (a, x ) + Γ(a, x ) .
2 ⎣⎢ ⎦⎥
2
Then, rearranging,
[n = 1,3,5,...] ⎛n ⎞
γ ⎜ , x ⎟ = Γ⎜ ⎟ − Γ⎜ , x ⎟
⎛n⎞ ⎛n ⎞
⎝2 ⎠ ⎝2⎠ ⎝2 ⎠
π
( ) π (n − 2 )!!
1
⎛3 ⎞ • ⎛n⎞
[n = 1,3,5, K] derived
γ ⎜ , x⎟ = Φ x − e−x x 2 Γ⎜ ⎟= n −1
⎝2 ⎠ 2 ⎝2⎠ 2 2
⎛5 ⎞ 3 π
( )
1
⎛3 ⎞ in Sect. 8.339, above
γ ⎜ , x⎟ = Φ x − e−x x 2 ⎜ + x ⎟
⎝2 ⎠ 4 ⎝2 ⎠ •
⎛1 ⎞
Γ⎜ , x ⎟ = π − π Φ x ( )
⎝2 ⎠
⎞ 15 π
( )
1
⎛7 ⎛ 15 5 ⎞
γ ⎜ , x⎟ = Φ x − e−x x ⎜ + x + x 2 ⎟
2
⎛n ⎞
⎝2 ⎠ 8 ⎝4 2 ⎠ • Use above result, Γ⎜ , x ⎟ =
⎝2 ⎠
⎡1 ⎛ 1 ⎞ ⎤
⎢ 2 Γ⎜ 2 , x ⎟ + ⎥
(n − 2)!! ⎢ ⎝ ⎠ ⎥
n −3 ⎢ n −3
1 2
(2 x ) s ⎥
2 2 ⎢e − x x 2 ∑ ⎥
⎢⎣ s = 0 (2 s + 1)(2 s − 1)!!⎥
⎦
=
(n − 2)!! π 1 − Φ x + [ ( )]
n −1
2 2

(2 x )s
n −3

(n − 2)!! e − x x 12 2

n −3 ∑ (2s + 1)(2s − 1)!!.

s =0
2 2

Thus,
⎛n
γ ⎜ , x⎟ =
⎞ (n − 2)!! π −
n −1
⎝2 ⎠
2 2

(n − 2)!! π + (n − 2)!! π Φ ( x )
n −1 n −1
2 2
2 2

(2 x )s
n −3

(n − 2 )!! − x 2
1 2

− e x
n −3 ∑ (2s + 1)(2s − 1)!!.
s =0
2 2

Factoring
(n − 2)!! ,
n −3
2 2

—Sect. 8.359—
 271

⎛n ⎞
γ ⎜ , x⎟ =
⎝2 ⎠
(n − 2)!! πΦ x −( )
n −1
2 2

(2 x )s
n −3

(n − 2)!! e − x x 12 2

n −3 ∑ (2s + 1)(2s − 1)!!

s =0
2 2

⎡1
⎢2 π Φ x − ( ) ⎤
⎥
=
(n − 2)!! ⎢
n −3
⎥
⎢ ⎥
(2 x )
n −3 1 2 2
2 2 ⎢e − x x 2 ∑ ⎥
⎢⎣ s =0 (2 s + 1)(2 s − 1)!!⎥⎦
which is the final form.

NOTE: if n = 1, γ ⎜ , x ⎟ = π Φ
⎛1 ⎞
( x ) by above
⎝2 ⎠
−1
since (−1)!!=1 and ∑ (•) = 0 , so formula is
0
general.
NOTE: Used in derivation of proposed formula,
u

∫ x e dx.
2 2
n −q u
3.321 8, above,
0
NOTE: as a check on calculations, add the results,
⎛n ⎞ ⎛n ⎞
Γ⎜ , x ⎟ + γ ⎜ , x ⎟ to get,
⎝2 ⎠ ⎝2 ⎠
⎛n⎞ π (n − 2 )!!
Γ⎜ ⎟= n −1
[n = 1,3,5, K]
⎝2⎠ 2 2

⎛n ⎞ • Derive as an exercise—follow steps in above,

γ ⎜ , x⎟
⎝k ⎠ ⎛n ⎞
Γ⎜ , x ⎟ based on relation,
⎡n n ⎤ ⎝k ⎠
⎢k > 0, ≠ 1,2,3, K⎥ Γ(α ) = γ (α , x ) + Γ(α , x )
⎣ k ⎦
• Compare result with Sect. 8.356, Form.
8.356.9
• Show that,
1
⎛7 ⎞ 4 ⎛1 ⎞ ⎛4 ⎞
γ ⎜ , x ⎟ = γ ⎜ , x ⎟ − x 3 e −x ⎜ + x ⎟
⎝3 ⎠ 9 ⎝3 ⎠ ⎝3 ⎠
• Show that,
⎛7⎞ ⎛7 ⎞ ⎛7 ⎞
Γ⎜ ⎟ = γ ⎜ , x ⎟ + Γ⎜ , x ⎟ or,
⎝3⎠ ⎝3 ⎠ ⎝3 ⎠
4 ⎛1⎞ 4 ⎡ ⎛1 ⎞ ⎛ 1 ⎞⎤
Γ⎜ ⎟ = ⎢γ ⎜ , x ⎟ + Γ⎜ , x ⎟⎥
9 ⎝ 3⎠ 9 ⎣ ⎝ 3 ⎠ ⎝ 3 ⎠⎦

—Sect. 8.359—
272
 273

Euler’s Constant γ or C

FORMULA PROOF OUTLINE

(formula references are to G & R 7e)
ln (ln t )
∞ ∞ ∞
ln t
C = −∫ dt In Formula 8.367.4, C = − ∫ e ln t dt = − ∫
−t
dt , use
1
t2 0 0
et
change of variable:
s = e t , ds = e t dt = sdt , ln s = ln e t = t , s0 = 1, s∞ = ∞, to
give,
ln(ln s )
∞
C = −∫ ds .
1
s2
• NOTE: submitted GR6 errata
¾ Verified math.com
∞ In Formula 8.367.11,
⎡ 1 1 ⎤ dt
C = ∫⎢ −
t − 1 t ln t ⎥⎦ t
∞
1 ⎣
⎡ 1 1⎤
C = ∫⎢ t − t ⎥ dt ,
0 ⎣
e − 1 te ⎦
set s = e t , ds = e t dt = sdt , ln s = t , s0 = 1, s∞ = ∞ , and
substitute to get
∞
⎡ 1 1 ⎤ ds
C = ∫⎢ −
1 ⎣
s − 1 s ln s ⎥⎦ s
¾ Verified math.com
∞
⎡ 1 1 ⎤ dt In Formula 8.367.9,
C = −∫ ⎢ −
t ln t ln t (1 + ln t ) ⎥⎦ t
∞ ∞
1 ⎣
⎡ −t 1 ⎤ dt ⎡1 1 ⎤ dt
C = − ∫ ⎢e − ⎥ = −∫ ⎢ t − ,
0 ⎣
1+ t ⎦ t 0 ⎣
e 1 + t ⎥⎦ t
set s = e t , ds = e t dt = sdt , ln s = t , s0 = 1, s∞ = ∞ , and
substitute to get
∞
⎡ 1 1 ⎤ ds
C = −∫ ⎢ −
1 ⎣
s ln s ln s(1 + ln s ) ⎥⎦ s
¾ Verified math.com

—Sect. 8.367—
 274

1
⎡1 1 ⎤
0
⎡ et et ⎤
C = ∫⎢ + • From below integral, C = ∫ ⎢ t et − 1⎥⎦ dt .
−
t ln (1 − t )⎥⎦
dt
0 ⎣ −∞ ⎣
1
⎡1 1 ⎤
• Substitute into hypothesis C = ∫ ⎢ +
t ln (1 − t ) ⎥⎦
dt ,
0⎣

s = ln (1 − t ), ds = −
1
dt , e s = e ln(1−t ) =
1− t
1 − t , s 0 = 0, s1 = −∞,
to get,
0
⎡ et et ⎤
C = ∫ ⎢ − t ⎥ dt .
−∞ ⎣
t e − 1⎦
¾ Verified math.com

∞ ∞
(
C = − ∫ t exp t − e dt t
) In Formula 8.367.4, C = − ∫ e −t ln tdt , use change of
−∞ 0

variable:
dt s
s = ln t , ds = , e = e ln t = t , s 0 =
t to give,
ln(0) = −∞, s ∞ = ∞,
∞ ∞
C = − ∫ se s e −e ds = − ∫ s exp(s − e s )ds .
s

−∞ −∞
• NOTE: submitted GR6 errata
¾ Verified math.com
0
⎡ et et ⎤ In Formula 8.367.6,
C= ∫ ⎢ t et − 1⎥⎦ dt
− 1
⎡ 1 1 ⎤
−∞ ⎣ C = ∫⎢ + dt ,
0⎣
ln t 1 − t ⎥⎦
dt s
set s = ln t , ds = , e = t , s0 = −∞, s1 = 0 , and substitute
t
to get
⎡es
0
es ⎤
C = ∫⎢ − s
s e − 1⎥⎦
ds
−∞ ⎣

¾ Verified math.com
∞ ¾ math.com calculation
C = ∫ te − (t + e )dt + Ei(−1)
−t

0
∞ ¾ math.com calculation
C = Ei(1) − ∫ te − (t − e )dt
−t

—Sect. 8.367—
 275

REFERENCES
Abramowitz, Milton and Irene Stegun. Handbook of Mathematical Functions with Formulas,
Graphs, and Mathematical Tables. Washington, DC: U.S.Government Printing Office, 1964
(Reprinted Dover, New York, 1972).

Artin, Emil. The Gamma Function. New York, NY: Holt, Rinehart and Winston, 1964.

Bers, Lipman. Calculus. 2 volumes. NY: Holt, 1969.

Boas, Mary L. Mathematical Methods in the Physical Sciences, 3 ed. NY: Wiley, 2006.

Boros, G. and Moll, V. Irresistible Integrals: Symbolics, Analysis and Experiments in the Evaluation
of Integrals. Cambridge, England: Cambridge University Press, 2004.

Carr, G.S. Formulas and Theorems in Pure Mathematics, 2 ed., New York: Chelsea, 1970

Dwight, H.B. Table of Integrals and Other Mathematical Data, 4th ed. New York: Macmillan, 1961.

Gautschi, W. “The incomplete gamma functions since Tricomi.” In Tricomi's Ideas and
Contemporary Applied Mathematics, Atti dei Convegni Lincei, n. 147, Accademia Nazionale dei
Lincei, Roma, 1998, pp. 203-237.

Gradshteyn, I.S. and I.M. Ryzhik (6th Edition). Table of Integrals, Series, and Products. Alan Jeffrey
and Daniel Zwillinger, Editors. NY: Academic Press, 2000.

_______ (7th Edition). Table of Integrals, Series, and Products. Alan Jeffrey and Daniel Zwillinger,
Editors. NY: Academic Press, 2007.

Moll, Victor H. The integrals in Gradshteyn and Rhyzik. Part 2: Elementary logarithmic integrals.
Scientia, Series A: Mathematical Sciences, Vol. 14 (2007), 7-15

_______. The integrals in Gradshteyn and Ryzhik. Part 3: Combinations of Logarithms and
Exponentials. arXiv:0705.0175v1 [math.CA]. 1 May 2007

_______.The integrals in Gradshteyn and Ryzhik. Part 4: The Gamma function

arXiv:0705.0179v1 [math.CA]. 1 May 2007.

_______. The integrals in Gradshteyn and Ryzhik. Part 13: Evaluation using The Error Function.
http://www.math.tulane.edu/~vhm/web_html/erfweb.pdf, 4 October 2006

_______. The integrals in Gradshteyn and Ryzhik. Part 14: The Exponential-Integral Function.
http://www.math.tulane.edu/~vhm/web_html/expintweb.pdf, 6 Semptember 2006.

_______. The integrals in Gradshteyn and Ryzhik. Part 15: The Incomplete Gamma Function
http://www.math.tulane.edu/~vhm/web_html/erfweb.pdf, 4 September 2006
 276

Moll, Victor H., Jason Rosenberg, Armin Straub, Pat Whitworth. The Integrals in Gradshteyn and
Ryzhik. Part 8: Combinations of powers, exponentials and logarithms. arXiv:0707.2123v1
[math.CA]. 14 Jul 2007.

O’Brien, F.J. 100 Proposed Formulas Submitted for Publication in I.S. Gradshteyn and I.M. Ryzhik
Table of Integrals, Series, and Products, 7th Edition. Alan Jeffrey and Daniel Zwillinger, Editors.
NY: Academic Press, 2005. Newport, RI: Naval Undersea Warfare Center, Division, Newport.
November 1, 2005.

_______. Selected Transformations, Identities, and Special Values for the Gamma Function.
Submittted to Mathworld, http://functions.wolfram.com/GammaBetaErf/Gamma, June, 2006.

_______. 400 Proposed Formulas Submitted for Publication in I.S. Gradshteyn and I.M. Ryzhik
Table of Integrals, Series, and Products, 8th Edition. Alan Jeffrey and Daniel Zwillinger, Editors.
NY: Academic Press, 2007. Newport, RI: Naval Undersea Warfare Center, Division, Newport.
May 9, 2007.

_______. Summary of Four Generalized Exponential Models (GEM) For Continuous Probability
Distributions, Jan. 18, 2008. arXiv:0801.2941v2 [math.GM]

Prudnikov, A. P., Yu. A. Brychkov, and O. I. Marichev. Integrals and Series. Vol. 1: Elementary
Functions. Gordon & Breach, New York, 1986.

_______. Integrals and Series. Vol. 2: Special Functions. Gordon & Breach, New York, 1986.

_______. Integrals and Series. Vol. 3: More Special Functions. Gordon & Breach, New York, 1990.

Sokolnikoff, I. S. Advanced Calculus. New York: McGraw Hill Book Co., 1939.

Spiegel, M. R. Mathematical Handbook of Formulas and Tables. New York: McGraw-Hill, 1968.
[Reprinted Spiegel, Murray R., John Liu, and Seymour Lipschutz. Mathematical Handbook of
Formulas and Tables. New York: McGraw-Hill, 1999.]

Whittaker, E. T. and G. N. Watson. A Course of Modern Analysis. Cambridge University Press; 4th
edition, 1934.

Wolfram Research, Inc. The Wolfram Functions Site. http://functions.wolfram.com/, 1998-2008.

_______. Wolfram Mathworld. http://mathworld.wolfram.com/, 2008.

Zwillinger, Daniel. Handbook of Integration. Boston: Jones & Bartlett Publishers, Inc., 1992.
 277

INDEX OF SYMBOLS, FUNCTIONS,

AND CONCEPTS
This index shows the occurrence of symbols, functions, and concepts used in the expressions
within the text. The numbers refer to pages on which the symbols, functions, or concepts appear.

(a )n see Pochhammer symbol li see logarithm - integral function

C or γ see Euler' s constant Li 2 see dilogarithm
G or K see Catalan' s constant lim see limit
ε see epsilon ln & log see logarithm
∉ not an element of a set n! see factorial
Ei see exponential - integral function Re real variable
erf see error functions Φ see error functions
erfi see error functions π mathematical constant 3.14159265...
exp or e see exponential function Π see product
i imaginary number Σ see summation
γ and Γ see gamma functions Ψ see Psi (Digamma) function
γ ∗ see gamma functions ⎛n⎞
⎜⎜ ⎟⎟ see binomial coefficient
int see integer function ⎝ j⎠
! and !! see factorials
 278

A
algebraic functions
and exponentials 9, 106-119
and logarithmic functions 62-77, 178-180
alternating series see oscillating series
arctan see arctangent
arctangent 17
arithmetic progression 25-26

B
binomial coefficient 20, 37-39,42-43, 45-46, 48-49, 50-51, 62, 65, 67-68, 73-76, 198

C
Catalan’s constant (G or K ) 17, 82, 180
Cauchy principal value (PV) see Principal Value
complement or reflection formula see gamma functions
complementary error function see error functions
complete gamma function see gamma functions
complex (imaginary) number i 2, 11, 33, 40, 47, 265-266
completing the square 20, 41, 89, 91
computer solution see mathematical software
constant of integration Preface, 128
cos cosine 229-234, 250
cosec see csc
csc cosecant 227-231

D
definite integrals 4-21, 81-275
derivatives see differentiation
differentiation 3, 5, 6, 8-9, 21, 102, 112-114, 264
of integrals 21, 34, 37, 42, 45, 81, 102, 114-115, 128, 264
of special functions 6, 8, 9, 21, 198, 263-264
digamma function see Psi function
dilogarithm Li 2 ( x ) 5, 55-56
double factorials see factorial, double
double integral see integral
 279

E
elementary functions 1-201
graphs 1-18
epsilon 12, 14
error functions 8-11, 33, 40-41, 44, 47, 49, 58-59, 85, 148-150, 214-229, 265-267
and gamma functions 11, 215-216
complementary 10, 215, 265-266
imaginary 10, 11, 33, 40-41, 47, 55, 269-272
proof of limits 58, 266-270
Euler’s constant (C or γ ) 6-7, 16, 101-102, 127, 178, 193, 198, 274-275
Euler’s integral see gamma functions
even function 59, 86, 122, 132, 267
exponential function 2-3
exponential-integral function 9, 12-14, 31-34, 36, 45-51, 57, 62-63, 65, 69, 71-77, 82, 99, 104, 152,
154, 163-164, 168-169, 170-171, 173, 193, 196, 198-201, 204-209, 211-213,
257-259, 265, 275
exponentials 2-3
and algebraic functions 9, 105-119
and complicated arguments 84-97, 129-159
and exponential integrals 209
and gamma functions 31-34, 37-60
and logarithmic functions 34, 36, 51, 77, 176
and powers 120-125, 129-159
and rational functions 35-60, 103-105, 126-128
of exponentials 98-102

F
factorial (!) 6, 20, 37-51, 62-63, 65, 67-68, 72-76, 87-88, 130-132, 146,165,
182-184, 239-240, 252, 256-257
factorial, double (!!) 20, 86, 87, 116-119, 130-133, 160, 182-189, 239-240, 243,
251, 268-272
finite Binomial expansion 20, 198
finite sums 24-26
Frullani Integral 81, 128

G
gamma functions 6-9
and complement or reflection formula 227-237
and exponentials 31-34
and logarithms 198
and multiplication theorem 227, 234, 238
and product theorem 227, 238
and trigonometric functions 225-238
complete (Γ(x)) 6-7, 225-254
doubling formula 229, 235, 238
 280

( )
incomplete γ ∗ ( x ) 255
integer arguments 86, 239-252
lower incomplete (γ(x, y)) 7-8, 56-60, 254-272
upper incomplete (Γ(x, y)) 7-9, 34, 36-60, 254-272
Gaussian probability density function 58-59, 86
geometric progression 26
graphs of elementary and special functions 1-18, 267

H
hypergeometric function 70

I
identities 7, 18-19
algebraic 18-19
trigonometric 7, 227-238
imaginary error function see error functions
imaginary number i see complex (imaginary) number
improper integrals 13, 15
incomplete gamma function see gamma functions
incorrect listing 212
indefinite integrals 4, 9, 11, 13-14, 27-77
integer function 22, 119, 244, 245-248
integral, double 59, 81, 128
inverse trigonometric functions see trigonometric functions

L
Leibnitz’s Rule for differentiating integrals see differentiation
L’Hôpital Rule 60
limit lim 2, 5, 12, 14, 16, 57, 60, 151, 159, 159, 255, 266-267
logarithm-integral function 12, 14-15, 31-32, 34, 62-66, 68-69, 71, 73-75, 82, 163-167, 137, 211, 212-
213
logarithms 4-5
and algebraic functions 62-78, 179-184
and dilogarithms 5, 55-56
and exponentials 77, 180, 195-201, 212
and gamma functions 198
and powers 62-78, 181-189
and rational functions 193-196
of more complicated arguments 177-178, 195-201

M
Math.com see quickmath.com
mathematical software Preface, also see Mathematica, quickmath.com
Mathematica Preface, 41, throughout text
 281

mathematical summary 1-21

Meijer’s G Function 264
multiplication theorem see gamma functions

N
natural logarithm see logarithms
natural numbers 6, 21, 118-119, 189, 225-226, 236-237, 242, 244-248
notations 22, also see Symbols

O
odd function 59
oscillating series 48, 56

P
partial fractions 18-19, 53-58, 111-113, 115
Pochhammer symbol 22, 225-226, 241, 242-252
polar coordinates 59
powers
and exponential functions 120-125, 129-159
and logarithmic functions 62-78, 181-191
principal value (PV) 12-15, 34, 57
probability integral Φ ( x ) see error function
product Π 22, 225, 238-252
product theorem see gamma functions
progressions 25-26
Psi function 6, 16

Q
quickmath.com Preface, throughout text

R
rational functions
and exponentials 103-105, 126-128
and logarithmic functions 192-193
references 3, 5, 7, 9 11, 13-17, 19-22, 58, 112, 298, 225-232, 238, 255, 276
reflection or complement formula see gamma functions

S
sec secant 234
series 5, 48, 55-57, 60, 65, 73, 75, 76
 282

sin sine 174, 232-238, 251

special functions 1-21, 202-275
graphs 10, 12, 14, 16-17, 267
Stirling asymptotic formula 20
summation Σ 18, 20, 22, 25-26, 37-40, 46, 48-51, 53-56, 57,59, 62-63, 65, 67-68, 73-76, 198, 256-259

T
transforms 18-19, 38, 43, 113, 238, 248, 264
transcendental functions 7, 9, 151
trigonometric functions 7, 227-238, also see arctangent, cosecant, cosine, secant, sine
 283

INDEX OF FORMULAS
This index lists the formulas which are solved in the text. The formulas are identified by the
section/subsection and page numbers where they appear in the text. Similar expressions are stated in
combined algebraic form. The Mathematical and Graphical Summary (page 1 and following) contains
other formulas not listed below.

FORMULA SECTION
int (x ) — NOTATIONS (22)
(a )n —
n —
∑u
k =m
k

n —
∏ f (k )
k =m
n −1 0.111 FINITE SUMS (25-26)
∑ (a + kr )
k =0
n 0.112
∑ aq
k =1
k −1

n 0.121.1
∑k
k =1
THE EXPONENTIAL FUNCTION (29-34)
∫x
m
e ± ax dx
n
2.31

e ± ax
n
2.31
∫ x m dx
∫ exp a dx
x
( ) 2.312

ax 2.312
∫ x a dx
ax 2.312
∫ x b dx
2x 2.312
∫ x 2 dx
∫ a dx
x2 2.312

dx 2.312
∫a x 2

∫ a dx
x n
2.312

dx 2.312
∫a x n

∫ x a dx
m xn 2.312

∫ e ln xdx THE EXPONENTIAL COMBINED WITH

μ ± x p 2.32
RATIONAL FUNCTIONS OF x (35-60)
∫ e ln xdx
μ ± x
 284

e ax 2.32
∫ x 4 dx
∫ x e dx
m ± ax n 2.32

∫ ln xdx 2.32
m

∫ x ln xdx 2.32
n m

∫ x ln xdx 2.32
n

erfi( x ) 2.32

∫e
ax 2
dx 2.32

( )dx
∫e
ax 2 +2 bx + c 2.32

Γ(a ), γ (a, x ), Γ(a, x ) 2.32

ln x n 2.32
∫ x m dx
erf ( x ) 2.32

∫x
m − βx n 2.32
e dx

∫x
a −1 − x 2.32
e dx
∫e
− βx 2
dx 2.32
n
e ax 2.32
∫ x m dx
n
e ax 2.32
∫ x dx
Ei(ax n ) 2.32
n
− βx n 2.32
e
∫ x m dx
e − βx
n
2.32
∫ x dx
e − βx
2
2.32
∫ x 2 dx
∫ x e dx 2.32
n ax

e ax 2.32
∫ x n dx
∫ x e dx
n − βx 2.32

∫ xe dx
− βx 2.32

∫ x e dx
β2 − x 2.32

∫ x e dx
β3 − x 2.32

e − βx dx 2.32
∫ x
 285

e − βx dx 2.32
∫ xn
e − βx dx 2.32
∫ x2
e − βx dx 2.32
∫ x3
e − βx dx 2.32
∫ x4
∫ x dx 2.32
n

1 2.32
∫ x n dx
1 1 2.32
∫ x 2 dx & ∫x 3
dx

∫ x dx 2.32
x

(− 1)1−γ
n
e ax γ −1 2.32
⎛ 1⎞
na γ ∫0
⎜ ln ⎟
⎝ t⎠
dt

e −1
ax
e ax + 1 2.32
∫ eax +1 ∫ eax −1 dx
dx &

e ax − m 2.32
∫ eax +n dx
e ax + m 2.32
∫ eax +n dx
e ax − m 2.32
∫ eax −n dx
∫ e dx 2.32
x

∫ e dx
−x 2.32

dx 2.32
∫ 1+ e ax

e ax 2.32
∫ x + bdx
x +1 2.32
∫ e ax dx
x 2.32
∫ e x − 1 dx
ln (a ± bx )dx 2.32
∫ x
ln ( x )dx 2.32
∫ a±x
 286

dx 2.32
∫ b
a+
e mx
dx 2.32
∫ b
a − mx
e
dx 2.32
∫ a − bemx
∫ x e dx
3 − βx 2 2.32

∫ x e dx
β
2 − x
2.32
Ei( x ) 2.32
li( x ) 2.32
xe x 2.32
∫ (1 + x )2 dx
xe ax 2.32
∫ (1 + ax )2 dx
x2
− 2.32
∫e 2
dx

Ei(ax ) 2.32
xn 2.32
lim =0
x → +∞ e x

∫ ln xdx 2.71 LOGARITHMS (61-77)

m

∫ ln(a + bx ) 2.71
m
dx
dx 2.72-2.73
∫ ln x
∫ x ln xdx 2.72-2.73

∫ x ln xdx 2.72-2.73
n m

∫ ln(ln x ) dx 2.72-2.73
n

∫ x ln(ln x ) dx 2.72-2.73
n m

ln (ln x ) 2.72-2.73
m

∫ x dx
xn 2.72-2.73
∫ (ln x )m dx

x n dx 2.72-2.73
∫ ln x 2
x n dx 2.72-2.73
∫ ln x 3
∫ (a + bx ) ln(c + ln x ) dx 2.72-2.73
m n
 287

ln(c + ln x ) 2.72-2.73
n

∫ a + bx dx
∫ (a + bx ) ln(c + kx ) dx 2.72-2.73
m n

ln(c + kx ) 2.72-2.73
n

∫ a + bx dx
ln x n 2.72-2.73
∫ (a + bx ) m
dx

ln x n dx 2.72-2.73
∫ a + bx
ln x n 2.72-2.73
∫ x m dx
ln x n 2.72-2.73
∫ x m dx
ln x 2.72-2.73
∫ x m dx
ln x 2 2.72-2.73
∫ x m dx
ln x 3 2.72-2.73
∫ x m dx
∫x ln(a + bx ) dx
n
m 2.72-2.73
ln (a ± bx )dx 2.72-2.73
∫ x
dx 2.72-2.73
∫ ln(a + bx )
x 2.72-2.73
∫ ln(a + bx ) dx
x n dx 2.72-2.73
∫ ln x
ln (a + ln x ) 2.72-2.73
m

∫ xn
dx

ln (a + ln x ) 2.72-2.73
m

∫ x
dx

∫ x ln(a + ln x ) dx 2.72-2.73
n m

∫ x ln(a + ln x )dx 2.72-2.73

n

ln (a + ln x ) 2.72-2.73
∫ x
dx

∫ x ln(a − ln x )dx 2.72-2.73

n

ln (a − ln x ) 2.72-2.73
∫ x
dx
dx 2.72-2.73
∫ (a − ln x ) n
 288

dx 2.72-2.73
∫ a − ln x
dx 2.72-2.73
∫x n
ln x m
dx 2.72-2.73
∫ x n ln x
dx 2.72-2.73
∫ (a + ln x )n
dx 2.72-2.73
∫ a + ln x
xn 2.72-2.73
∫ (a + ln x ) m
dx

xn 2.72-2.73
∫ a − ln x dx
ln x 2.72-2.73
∫ (a ± ln x )2 dx
n
e ax 2.72-2.73
∫ x m dx
∫ x e dx
m ax n 2.72-2.73
v 3.310 EXPONENTIAL FUNCTIONS (81-83)
∫ e dx
−ρ x

u
∞
e − x − e −a x 3.310
∫0 x dx
∞ 3.310
∫
f ( ax )− f ( bx )
dx
x
0
∞
e − ax − e − bx
p p
3.310
∫0 x
dx

0
xe x 3.311
−∫ dx
− ∞1 + e
2x

n
u
ex ln a 3.311
∫
−∞
x
dx

v 3.311
∫ a dx
n
x

u
1 3.311
∫ a dx
n
x

0
1
1 3.311
∫a
0
x n
dx

∞ 3.311
1
∫a
0
xn
dx
 289

1 3.311
∫ a dx
x

0
1 3.311
∫ a dx
n
x

0
u 3.321 EXPONENTIALS OF MORE COMPLICATED
∫ e dx
−x 2

ARGUMENTS (84-97)
0

2
u 3.321
∫ e dx
−x 2

π 0
v 3.321
∫ e dx
−q x 2 2

u
∞ 3.321
∫ e dx
−q x 2 2

u
∞ 3.321
∫ e dx
−q x 2 2

−∞
u 3.321
∫ x e dx
n −q x 2 2

0
v
⎛ x2 ⎞ 3.322
∫u ⎜⎜⎝ − 4β − γx ⎟⎟⎠dx
exp

∞ 3.323
(
∫ exp − ax − bx dx
2
)
0
∞ 3.323
(
∫ exp − ax − bx dx
2
)
u
3.323
( )
u

∫ exp − ax − bx dx
2

0
v
(ax 2
+ kbx + c ) 3.323
∫e
−
dx
u
v
(ax 2
+ 2 bx + c ) 3.323
∫e
−
dx
u
∞
(ax 2
+ kbx + c ) 3.323
∫e
−
dx
0
∞
(ax 2
+ bx + c ) 3.323
∫e
−
dx
0
∞
(ax 2
+ bx + c ) 3.323
∫e
−
dx
u
u
(ax 2
+ kbx + c ) 3.323
∫e
−
dx
0
 290

∞
(ax ) 3.323
∫ (cx )
2
+ jbx
2
+ kgx e − dx
−∞
∞
(ax )
∫ (cx )
2
+ bx
3.323
2
+ gx e − dx
0
∞
( px )
∫ (ax )
2
+ kqx 3.323
2
+ bx + c e − dx
−∞
∞
( px )
∫ (ax )
2
+ qx + c 3.323
2
+ kbx e − dx
−∞
∞
(∫ ax + kbx+ c)e ( ) 3.323
− px + jqx+r
2
2
dx
−∞

⎛ a+b⎞ 3.324
u

∫0 exp⎜⎝ − x 2 ⎟⎠dx
( )
v 3.326
∫ x exp − βx dx
m n

u
v
exp − β x n ( ) 3.326
∫u x m dx
∞
( )
3.326
∫ x exp − βx dx
m n

0
v 3.326
∫ x exp(− ρx − b )dx
m

u
∞
( )
3.326
∫ x exp − βx dx
m + n +1 n +1

0
∞ 3.326
∫ ( x − a )e
− β ( x −b) n
dx
0
∞ 3.326
∫ ( x − a )e
− β ( x −b) n
dx
u

∞ 3.326
∫ ( x + a )e
− β ( x −b) n
dx
0

∞ 3.326
∫ ( x + a )e
− β ( x −b) n
dx
u
u 3.326
∫ ( x + a )e
− β ( x −b) n
dx
0

3.327 EXPONENTIALS OF
( )
v

∫ exp − ae dx
nx
EXPONENTIALS (99-102)
u
 291

∞ 3.327
− (x + e − x )
∫ xe
0
dx

∞ 3.327
− (x − e − x )
∫ xe
0
dx

3.328
( )
v

∫ exp − e e dx
x μx

u
3.328
( )
v

∫ exp − e e dx
−x μx

u
∞
( )
3.331
∫ x exp x − e dx
x

−∞
3.331
( )
v

∫ exp − βe − μx dx
−x

u
3.331
( )
v

∫ exp − βe − μx dx
x

u
3.331
( )
v

∫ exp − βe + μx dx
x

u
∞

∫ exp(− βe ) 3.331
x
+ μx dx
0

3.331
( )
v

∫ exp − βe + μx dx
−x

u
∞
β μ ∫ x μ −1e − βx dx
3.331
1

1 3.331
∫ x e dx
μ μ −1 − βx
β
0
∞ 3.331
β μ ∫ x μ −1e −βx dx
0

3.331
γ (μ , β )
d
dβ
3.331
Γ(μ , β )
d
dβ
v
e − ρx 3.351 COMBINATIONS OF EXPONENTIALS AND
∫u x n+1 dx RATIONAL FUNCTIONS (103-105)
∞
e−x 3.351
−∫ dx
−u
x
u 3.351
∫ xe dx
− μx

0
u 3.351
∫ x e dx
2 − μx

0
 292

u 3.351
∫ x e dx
3 − μx

0
v 3.351
∫ x e dx
n − μx

u
v
xe x 3.353
∫ (1 + x )
u
2
dx

xe − x 3.353
v

∫ (1 − x )
u
2
dx

v
e − qx 3.361 COMBINATIONS OF EXPONENTIALS AND
∫
u x
dx ALGEBRAIC FUNCTIONS (106-118)
∞
e − qx 3.361
∫
u x
dx

e − ρx 3.362
v

∫
u ax ± b
dx

∞
e − μx 3.362
∫
u x+β
dx

e − μx 3.362
u

∫
0 x+β
dx

∞
e − μx 3.362
∫
u x−u
dx

∞ 3.363
∫ x x − u e dx
− μx

u
v 3.363
∫ x a + bxe dx
− px

u
v 3.363
∫ x(a + bx ) e − px dx
m

u
∞ 3.363
x
∫
u x−u
e − μx dx

∞ 3.363
x − px
∫0 a + xe dx
v
x 3.363
∫u (a + bx )m e dx
− px

∞
ae − px
2
3.363
∫0 a + x 2 dx
∞
x − u − qx 3.363
∫
u
x
e dx
 293

∞ 1 3.371
n−
∫x
0
2
e − μx dx

∞ 1 3.371
n+
∫x
0
2
e − μx dx

∞ 1 3.371
−n
∫ x 2 e dx
− μx

0
∞ n−
p 3.371
∫x e − μx dx
q

0
∞ n+
p 3.371
∫x e − μx dx
q

0
∞ p
−n 3.371
∫x e − μx dx
q

0
u 3.381 COMBINATIONS OF EXPONENTIALS AND
∫ x e dx
m − βx n

ARBITRARY POWERS (120-125)

0
∞ 3.381
∫ x e dx
m − βx n

u
∞ 3.381
∫ x e dx
m − βx n

0
u 3.381
∫ x e dx
ρ −1 − x

0
∞
e−x 3.381
∫u xν dx
v 3.381
∫ x e dx
ν −1 − μx

∫
∞

−∞
(
x 2 m exp − β x 2 n dx) 3.381

∞ 3.381
∫x e
n±b
m ± a − βx

0
u 3.381
∫x e
n±b
m ± a − βx

0
∞ 3.381
∫x e
n±b
m ± a − βx

u
∞ 3.382
∫ (1 + x ) e dx
−ν − μx

0
0
⎡1 1 ⎤ x 3.427 COMBINATIONS OF RATIONAL FUNCTIONS
∫−∞⎢⎣ x − e x − 1⎥⎦e dx OF POWERS AND EXPONENTIALS (126-128)
 294

∞
e − vx − e − μx
t t
3.434
∫0 x ρ +1 dx
exp(− ax ) − exp(− bx )dx
∞ 3.434
∫0 x
∞ 3.461 COMBINATIONS OF EXPONENTIALS OF
∫x
n −1 − ρx 2
e dx MORE COMPLICATED ARGUMENTS AND
0 POWERS (129-160)
∞ 3.461
∫x
2 n −1 − ρx 2
e dx
0
∞ 3.461
2 ( n −1) − ρx
∫ x e dx
2

0
∞
e−ρ x
2 2
3.461
∫0 x 2n dx
∞ 3.461
∫ e dx
− ρx 2

−∞
∞ 2 3.461
2 n − ρx
∫x e dx
−∞
v
− qx
2 dx 3.461
∫e
u x2
v 3.462
∫ x e dx
n ±b
m± a − βx

u
∞ 3.462
∫ e dx
n±a
− βx

0
u 3.462
∫ e dx
n±a
− βx

0
∞ 3.462
∫ e dx
n±a
− βx

u
∞ 3.462
∫ (x ± b )e
−βx n ± b
dx
0
v 3.462
∫ (x ± b )e
n ±b
− βx
dx
u
n ±b
e − βx 3.462
v

∫u x m±a dx
∞ n ±b
e − βx 3.462
∫0 x m±a dx
∞ n ±b
e − βx 3.462
∫u x m±a dx
 295

n ±b
e − βx 3.462
u

∫0 x m±a dx
v 3.462
∫ (ax ± b ) e dx
m − ρx

u
v 3.462
∫ (b − ax ) e − ρx dx
m

u
∞ 3.462
∫ (b − ax ) e dx
m − ρx

0
u 3.462
∫ (b − ax ) e dx
m − ρx

0
∞ 3.462
∫ (b − ax ) e dx
m − ρx

e − ρx 3.462
v

∫u (ax ± b )n dx
e − ρx 3.462
v

∫u (b − ax )n dx
∞
e − ρx 3.462
∫0 (b − ax )n dx
e − ρx 3.462
u

∫0 (b − ax )n dx
∞
e − ρx 3.462
∫u (b − ax )n dx
n±q
⎛ x−a ⎞
−β ⎜ ⎟
3.462
v ⎝ b ⎠
e
∫ ( )
u
x−a m± p dx
b
∞ 3.462
∫ (x − a )e
−β ( x± a )
dx
0
∞ 3.462
∫ (ax ± b ) e dx
m − px

0
∞ 3.462
∫ (ax ± b ) e dx
m − px

3.462
u

∫ (ax ± b ) e − px dx
m

∞
e − px 3.462
∫0 (ax ± b )n dx
 296

∞
e − px 3.462
∫u (ax ± b )n dx
e − px 3.462
u

∫0 (ax ± b )n dx
∞ 3.462
∫ (x + a )e
− β ( x− a )
dx
0
n± q
∞ m± p ⎛ x−a ⎞ 3.462
⎛ x−a⎞ −β ⎜ ⎟

∫0 ⎜⎝ b ⎟⎠ e ⎝ b ⎠
dx
n± q
m± p ⎛ x−a ⎞ 3.462
⎛ x−a⎞
u −β ⎜ ⎟

∫0 ⎜⎝ b ⎟⎠ e ⎝ b ⎠
dx
n± q
∞ m± p ⎛ x−a ⎞ 3.462
⎛ x−a⎞ −β ⎜ ⎟

∫u ⎜⎝ b ⎟⎠ e ⎝ b ⎠
dx

∞ 3.462
⎛x−a⎞ ⎛x−a⎞
j k

∫0 ⎜⎝ b ⎟⎠ exp β ⎜⎝ b ⎟⎠ dx
∞
e − βx
n
3.462
∫0 x m dx
e − βx
n
u 3.462
∫0 x m dx
∞
e − βx
n
3.462
∫u x m dx

∫ (e ) dxx
∞ 3.464
− μx 2
− e − βx
2

2
u
u
⎛ − μx 2 − βx 2 ⎞ dx 3.464
∫ ⎜e −e ⎟ 2
0⎝ ⎠x
v
⎛ b ⎞ 3.471
∫ exp⎜⎝ − x
u
n ⎟
⎠
dx

u
⎛ b ⎞ 3.471
∫0 exp⎜⎝ − x n ⎟⎠dx
u
⎛ b⎞ 3.471
∫0 exp⎜⎝ − x ⎟⎠dx
u
⎛ b ⎞ 3.471
∫0 exp⎜⎝ − x 2 ⎟⎠dx
⎛ β ⎞ dx 3.471
u

∫0 exp⎜⎝ − x n ⎟⎠ x n+1
[ ]dx
v 3.471
∫ exp − β (ax + b )
−n

u
 297

[ ]dx
u 3.471
∫ exp − β (ax + b )
−n

[ ]
u 3.471
∫ exp − β (ax + b ) dx
−1

∫ exp[− β (ax + b ) ]dx

u
−2
3.471

0
u
⎡ β (ax + b ) ⎤ 3.471
∫0 exp⎢⎣− ax + b ⎥ dx
⎦

[ ]dx
v 3.471
∫ exp − β (b − ax )
−n

[ ]dx
u 3.471
∫ exp − β (b − ax )
−n

∫ exp[− β (b − ax ) ]dx
u
−1
3.471

[ ]dx
u 3.471
∫ exp − β (b − ax )
−2

0
u
⎡ β ⎤ 3.471
∫0 exp⎢⎣− b − ax ⎥⎦dx
u
⎡ b ⎤ 3.471
∫0 exp⎢⎣− a + x ⎥⎦ dx
[ ( ) ]dx 3.471
u
−1
∫ exp − β x
n+ p

⎛ b ⎞ 3.471
u exp⎜ − n ⎟
⎝ x ⎠ dx
∫0 x m
β 3.471
−
v xn
e
∫u x m dx
β 3.471
−
∞ n
e x
∫0 x m dx
β 3.471
−
∞ xn
e
∫u x m dx
⎛ b ⎞ 3.471
u

∫x exp⎜ − n ⎟dx
m

0 ⎝ x ⎠
 298

v
(a − x )μ −1 exp⎛ − β ⎞dx 3.471
∫
u x μ +1
⎜
⎝
⎟
x⎠
∞

∫ exp(− x )x
n( m +1 2 )n −1
dx
3.473
0

∞
( )
3.473
( m −1 2 )n −1
∫ exp − βx x
n
dx
0
∞

∫ exp(− βx )x
3.473
n (1 2− m )n −1
dx
0
v dx 4.211 LOGARITHMIC FUNCTIONS (161-178)
∫
u ± ln x
∞ 4.211
dx
∫e ln x
v 4.211
∫ ln xdx
u
v
dx 4.211
∫u x ln x
v
x n dx 4.211
∫u ln x
u
x n dx 4.211
∫0 ln x
v
xdx 4.211
∫u ln x
v 4.211
∫ x ln xdx
u
e 4.211
∫ ln x dx
1
v 4.211
∫ ln(ln x )dx
u
v
(ln x ) p dx 4.211
∫
u
x
e
ln xdx 4.211
∫1 x
e
(ln x ) p dx 4.211
∫
1
x
1
ln xdx 4.212
∫ a + ln x
0
 299

1
ln xdx 4.212
∫0 a − ln x
v
dx 4.212
∫u (a ± ln x )2
v
dx 4.212
∫ (a ± ln x )
u
n

v
ln xdx 4.212
∫u (a + ln x )n
v
ln xdx 4.212
∫u (a − ln x )n
v
ln xdx 4.212
∫u (a ± ln x )2
μ −1 4.215
⎛ 1⎞
1

∫0 ⎜⎝ ln x ⎟⎠ dx

1
dx 4.215
∫⎛ 1⎞
μ

⎜ ln ⎟
0

⎝ x⎠
n
−1 4.215
⎛ 1 ⎞2
1

∫0 ⎜⎝ ln x ⎟⎠ dx
1
dx 4.215
∫ n
+1
0
⎛ 1⎞ 2
⎜ ln ⎟
⎝ x⎠
μ −1 4.215
⎛ 1⎞
v

∫u ⎜⎝ ln x ⎟⎠ dx

γ −1 4.215
⎛ 1⎞
1
1
γ ∫⎜
ln ⎟ dt
nβ 0 ⎝ t ⎠
βz 1
dt 4.215
n ∫⎛ 1⎞
z +1

⎜ ln ⎟
0

⎝ t⎠
ln (ln x ) LOGARITHMS OF MORE
∞ 4.229
∫1 x 2 dx COMPLICATED ARGUMENTS (178)

1
∞
ln x 4.241 COMBINATIONS OF LOGARITHMS AND
∫
4 1 x ( x + 1)
dx ALGEBRAIC FUNCTIONS (179-180)
n 4.269 COMBINATIONS INVOLVING POWERS OF
⎛ 1 ⎞ 2 p −1
1

∫0 ⎜⎝ ln x ⎟⎠ x dx THE LOGARITHM AND OTHER POWERS

(181-191)
 300

x p −1 4.269
1

∫ n
dx
0
⎛ 1 ⎞2
⎜ ln ⎟
⎝ x⎠
1 4.272
∫ (ln x ) x dx
n p

0
1 4.272
∫ (ln x )x dx
p

0
1
−n 4.272
⎛ 1 ⎞ 2 ν −1
1

∫0 ⎜⎝ ln x ⎟⎠ x dx
n+
1 4.272
⎛ 1⎞
1
2
∫0 ⎜⎝ ln x ⎟⎠
ν −1
x dx

∞
(ln x ) p ± a ± n−1 dx 4.272
∫
1 x2
∞
(ln x ) p + a −1 dx 4.272
∫
1 x2
∞
(ln x ) p −1 dx 4.272
∫
1 x2
v
(ln x ) p −1 dx 4.272
∫
u x 2

∞ 4.272
dx
∫1 (ln x ) p x 2
∞ 4.272
dx
∫ (ln x )
1
p +1
x2
1 4.274
e q
x
∫ n
dx
x[− (1 + ln x )]
2
0

∞
⎡ 1 1 ⎤ dx 4.281 COMBINATIONS OF RATIONAL FUNCTIONS
∫ ⎢⎣ x − 1 − x ln x ⎥⎦ x
OF ln x AND POWERS
1 (193-194)
∞ 4.281
dx
∫ x (ln p + ln x )
1
2

x n + a −1 4.281
u

∫0 ln x dx
x n −1 4.281
u

∫0 ln x dx
∞
⎡ 1 1 ⎤ dx 4.283
∫1 ⎢⎣ x ln x − ln x(1 + ln x )⎥⎦ x
 301

1
⎡1 1 ⎤ 4.283
∫ ⎢⎣ x + ln(1 − x )⎥⎦ dx
0
1 4.326 COMBINATIONS OF LOGARITHMIC
∫ ln(a + ln x )x dx
μ −1
FUNCTIONS OF MORE COMPLICATED
0 ARGUMENTS AND POWERS (196)

∞ 4.331 COMBINATIONS OF LOGARITHMS

∫ e (ln x ) dx
− μx n
AND EXPONENTIALS (197-201)
0
v 4.331
∫ e ln xdx
−μ x

u
v 4.331
∫ e ln xdx
μx

u
v 4.337
∫ e ln(a ± bx )dx
−μ x

u
∞ 4.337
∫ e ln(a + bx )dx
−μ x

0
∞ 4.337
∫e
−μ x
ln(a ± bx )dx
u
∞ 4.337
∫ e ln(a ± bx )dx
−μ x

± Γ(0,± x ) 8.212 THE EXPONENTIAL INTEGRAL

FUNCTION Ei(x) (203-209)
(
Ei ± x n ) 8.212
Ei(± x ) 8.212
⎛ 1 ⎞ 8.212
Ei⎜ ± n ⎟
⎝ x ⎠
⎛ 1⎞ 8.212
Ei⎜ ± ⎟
⎝ x⎠
Ei( x ± y ) 8.212
Ei[− ( x + y )] 8.212
⎛ 1 ⎞ 8.212
Ei⎜⎜ ⎟⎟
⎝x± y⎠
⎛ x⎞ 8.212
Ei⎜⎜ ± ⎟⎟
⎝ y⎠
Ei a x ( ) 8.212
li(x ) 8.240 THE LOGARITHM INTEGRAL li(x) (211)
− Γ(0,− ln x ) 8.240
ln x
et 8.241 INTEGRAL REPRESENTATIONS (212-213)
∫ t dt
−∞
 302

( )
li a x 8.241
li( xy ) 8.241
erfc( x ) 8.250 THE PROBABILITY INTEGRAL, THE
FRESNEL INTEGRALS Φ ( x ), S ( x), C ( x) , THE
ERROR FUNCTION ERF(x),
AND THE COMPLEMENTARY ERROR
FUNCTION ERFC( x ) (214-220)
Φ ( xy ) 8.252
⎛ x⎞ 8.252
Φ⎜⎜ ⎟⎟
⎝ y⎠
⎛ y ⎞ 8.252
Φ⎜ ⎟
⎝ 2x ⎠
⎛x± y⎞ 8.252
Φ⎜ ⎟
⎝ a ⎠
Φ(x ± y ) 8.252
⎛x± y⎞ 8.252
Φ⎜ ⎟
⎝ 2 ⎠
Φ (a + bx ) 8.252
Φ ax ( ) 8.252
⎛1− z ⎞ 8.313 THE GAMMA FUNCTION (EULER’S
Γ⎜ ⎟
⎝ v ⎠ INTEGRAL OF THE SECOND KIND): Γ(z)
(221)
⎛ z −1⎞ 8.313
Γ⎜ ⎟
⎝ v ⎠
Γ(x ± a ) 8.331 FUNCTIONAL RELATIONS INVOLVING THE
GAMMA FUNCTION (222-238)
Γ(x ± n ) 8.331
Γ(n − x ) 8.331
(n − x )Γ(n − x ) 8.331
Γ(1 − x )
Γ(− x ) 8.331
⎛ x⎞ 8.331
Γ⎜ ± ⎟
⎝ 2⎠
⎛ x⎞ 8.331
Γ⎜1 − ⎟
⎝ 2⎠
⎛1− x ⎞ 8.331
Γ⎜ ⎟
⎝ 2 ⎠
Γ(m )Γ(1 − m ) 8.331
⎛ x ± 1⎞ 8.331
Γ⎜ ⎟
⎝ 2 ⎠
Γ(x − 1) 8.331
 303

⎛1 ⎞ 8.331
Γ⎜ − x ⎟
⎝2 ⎠
Γ(x )Γ(− x ) 8.334
Γ(1 + x )Γ(1 − x ) 8.334
Γ(1 + x )Γ(− x ) 8.334
Γ(− x )Γ( x − 1) 8.334
⎛ 1⎞ 8.334
Γ( x )Γ⎜ x + ⎟
⎝ 2⎠
⎛ 1⎞ 8.334
Γ⎜ x − ⎟
⎝ 2⎠
⎛ 1⎞ ⎛1 ⎞ 8.334
Γ ⎜ x − ⎟ Γ⎜ − x ⎟
⎝ 2⎠ ⎝2 ⎠
⎛ 1⎞ ⎛ 1⎞ 8.334
Γ⎜ x + ⎟Γ⎜ x − ⎟
⎝ 2⎠ ⎝ 2⎠
Γ(2 x ) 8.334
⎛ x⎞ ⎛ x ⎞ 8.334
Γ⎜1 − ⎟Γ⎜ − 1⎟
⎝ 2⎠ ⎝2 ⎠
Γ(n ± x )Γ( x − n ) 8.334
Γ(1 + x )Γ( x − 1) 8.334
Γ(n − x )Γ( x + n ) 8.334
Γ(x )Γ(1 − x ) 8.334
Γ(x )Γ(n − x ) 8.334
sin πx 8.335
2 n −1 8.335
n
8.335
Γ(x ) &
1
definitions
Γ( x)
n −1 ⎛ k⎞ 8.335
∏ Γ⎜ z + ⎟
k =0 ⎝ n⎠
Γ(2 z ), Γ(3 z ), Γ(nz ) 8.335
Γ(nz + b ) 8.335
n −1
⎛k⎞ 8.335
∏ Γ⎜ ⎟
k =1 ⎝ n ⎠
n −1
⎛ k ⎞⎛ k ⎞ 8.335
∏ Γ⎜ ⎟⎜1 − ⎟
k =1 ⎝ n ⎠⎝ n⎠
Γ(n ± k ) 8.339 PARTICULAR VALUES: FOR N A NATURAL
NUMBER (239-252)
Γ(n + 1) 8.339
⎛ n⎞ 8.339
Γ⎜ ± ⎟
⎝ 2⎠
 304

⎛n ⎞ 8.339
Γ⎜ + 1⎟
⎝2 ⎠
⎛ p⎞ 8.339
Γ⎜ n ± ⎟
⎝ q⎠
⎛ 1⎞ ⎛ 3⎞ 8.339
Γ⎜ n ± ⎟ ,…, Γ⎜ n ± ⎟
⎝ 2⎠ ⎝ 4⎠
⎛p ⎞ 8.339
Γ⎜ − n ⎟
⎝q ⎠
⎛1 ⎞ ⎛3 ⎞ 8.339
Γ⎜ − n ⎟ ,…, Γ⎜ − n ⎟
⎝2 ⎠ ⎝4 ⎠
⎛ p⎞
Γ⎜ ± ⎟
8.339
⎝ q⎠
⎛ 1⎞ ⎛1 ⎞ 8.339
Γ⎜ n ± ⎟Γ⎜ − n ⎟
⎝ 2⎠ ⎝2 ⎠
⎛n⎞ ⎛ n⎞ 8.339
Γ ⎜ ⎟Γ ⎜ − ⎟
⎝2⎠ ⎝ 2⎠
Γ(n + x ) 8.339
Γ( x − n )
Γ(z ,0 ) 8.350 THE INCOMPLETE GAMMA FUNCTION
DEFINITION (253-255)
Γ(a, ∞ ) 8.350
γ (a,0) 8.350
γ (a, ∞ ) 8.350
γ (− n, x )
* 8.351
γ (n ± k , x ) 8.352 SPECIAL CASES (256-260)
Γ(n ± k , x ) 8.352
Γ(− n + k , x ) 8.352
Γ(0, x ), Γ(1, x )Γ(2, x ), Γ(3, x ), 8.352
Γ(4, x )
Γ(1,− x ), Γ(2,− x ), Γ(3,− x ), Γ(4,− x ) 8.352

Γ(− 1, x ), Γ(− 2, x ), Γ(− 3, x ), Γ(− 4, x ) 8.352

Γ(0,− x ), Γ(− 1,− x ), Γ(− 2,− x ), Γ(− 3,− x ), 8.352
Γ(− 4,− x )

γ (1, x ), γ (2, x ), γ (3, x ), γ (4, x ) 8.352

γ (1,− x ), γ (2,− x ), γ (3,− x ), γ (4,− x ) 8.352

γ (α , x ) 8.353 INTEGRAL REPRESENTATIONS (261-262)
γ (α , x ± y ) 8.353
γ (α , xy )
⎛ x⎞ 8.353
γ ⎜⎜ α , ⎟⎟
⎝ y⎠
 305

Γ(α , x ) 8.353
⎛ x⎞ 8.353
Γ⎜⎜ α , ⎟⎟
⎝ y⎠
Γ(α , x ± y ) 8.353
Γ(a ± k , x ) 8.356 FUNCTIONAL RELATIONS (263-264)
γ (a ± k , x ) 8.356
dΓ(a, x ) 8.356
da
dγ (a, x ) 8.356
da
8.356
erf ( x )
d
dx
[Ei(x )] & d [li(x )]
d 8.356
dx dx
dγ (a, x ) 8.356
dx
dΓ( x ) 8.356
dx
d ⎛ 1 ⎞ 8.356
⎜ ⎟
dx ⎜⎝ Γ( x ) ⎟⎠
x 8.356
exp( xy )dy `
d
dx ∫0
1 (
d γ ν , βu n ) 8.356
nβ ν du nβ ν
1 (
d Γ γ , βu n ) 8.356
nβ γ du nβ γ
(
d Γ γ , βu n ) 8.356
dγ nβ γ
(
d γ ν , βu n ) 8.356
dν Γ(ν )
⎛1 ⎞ ⎛1 ⎞ 8.359 RELATIONSHIPS WITH OTHER FUNCTIONS
γ ⎜ ,± x k ⎟, Γ⎜ ,± x k ⎟ (265-272)
⎝2 ⎠ ⎝2 ⎠
⎡1
γ ⎢ , ax ⎥( )
2⎤ 8.359
⎣2 ⎦
⎛ 1 ⎞ 8.359
Γ⎜ ± , ± x ⎟
⎝ 2 ⎠
⎛ 1 ⎞ 8.359
Γ⎜ ± , ± x 2 ⎟
⎝ 2 ⎠
⎛1 ⎞ 8.359
γ ⎜ ,±x ⎟
⎝2 ⎠
 306

⎛1 ⎞ 8.359
γ ⎜ ,± x 2 ⎟
⎝2 ⎠
1 ⎡ ⎛1 2⎞ ⎛ 1 2 ⎞⎤ 8.359
⎢γ ⎜ 2 , x ⎟ + Γ⎜ 2 , x ⎟⎥
π ⎣ ⎝ ⎠ ⎝ ⎠⎦
erf (± ∞ ) 8.359
⎛ n ⎞ 8.359
Γ⎜ ± , ± x ⎟
⎝ 2 ⎠
⎛n ⎞ 8.359
Γ⎜ , x ⎟
⎝k ⎠
⎛n ⎞ 8.359
γ ⎜ , x⎟
⎝k ⎠
ln (ln t ) THE PSI FUNCTION. EULER’S CONSTANT:
∞ 8.367
−∫ dt INTEGRAL REPRESENTATIONS (273-275)
1 t2
∞ 8.367
⎡ 1 1 ⎤ dt
∫ ⎢⎣ t − 1 − t ln t ⎥⎦
1
t
∞
⎡ 1 1 ⎤ dt 8.367
−∫⎢ −
1 ⎣
t ln t ln t (1 + ln t )⎥⎦ t
1
⎡1 1 ⎤ 8.367
∫0 ⎢⎣ t + ln(1 − t )⎥⎦ dt
∞
( )
8.367
− ∫ t exp t − e t dt
−∞
0
⎡ et et ⎤ 8.367
∫ ⎢ t e t − 1⎥⎦ dt
− ∞⎣
−

∞ 8.367
− (t + e − t )
∫ te dt + Ei(−1)
0
∞ 8.367
Ei(1) − ∫ te −(t −e )dt
−t

0
 About The Author

Dr. Francis (Frank) J. O’Brien, Jr. received his Ph.D. from Columbia. He is a Senior Scientist
with the Department of Navy and works in the Undersea Warfare Combat Systems Department
at the Naval Undersea Warfare Center, Newport, Rhode Island. His areas of expertise are
statistical signal processing and logistics modeling. O’Brien has contributed extensively to
Gradshteyn and Ryzhik’s Table of Integrals, Series and Products. He holds numerous U.S.
Patents involving engineering mathematics.