Order of integration (calculus)

For the summary statistic in time series, see Order of y y

integration.
z z
In calculus, interchange of the order of integration
is a methodology that transforms iterated integrals (or dy
multiple integrals through the use of Fubinis theorem) a a
dx
of functions into other, hopefully simpler, integrals by
changing the order in which the integrations are per- a z x a z x
formed. In some cases, the order of integration can be
validly interchanged; in others it cannot. Figure 1: Integration over the triangular area can be done using
vertical or horizontal strips as the rst step. This is an overhead
view, looking down the z-axis onto the x-y plane. The sloped line
1 Problem statement is the curve y = x.

The problem for examination is evaluation of an integral In this expression, the second integral is calculated rst
of the form with respect to y and x is held constanta strip of width
dx is integrated rst over the y-direction (a strip of width
dx in the x direction is integrated with respect to the
y variable across the y direction), adding up an innite
f (x, y) dx dy,
D amount of rectangles of width dy along the y-axis. This
forms a three dimensional slice dx wide along the x-axis,
where D is some two-dimensional area in the xyplane.
from y=a to y=x along the y axis, and in the z direction
For some functions f straightforward integration is feasi-
z=f(x,y). Notice that if the thickness dx is innitesimal,
ble, but where that is not true, the integral can sometimes
x varies only innitesimally on the slice. We can assume
be reduced to simpler form by changing the order of inte-
that x is constant.[3] This integration is as shown in the left
gration. The diculty with this interchange is determin-
panel of Figure 1, but is inconvenient especially when the
ing the change in description of the domain D.
function h ( y ) is not easily integrated. The integral can
The method also is applicable to other multiple inte- be reduced to a single integration by reversing the order
grals.[1][2] of integration as shown in the right panel of the gure.
Sometimes, even though a full evaluation is dicult, or To accomplish this interchange of variables, the strip of
perhaps requires a numerical integration, a double inte- width dy is rst integrated from the line x = y to the limit
gral can be reduced to a single integration, as illustrated x = z, and then the result is integrated from y = a to y =
next. Reduction to a single integration makes a numerical z, resulting in:
evaluation much easier and more ecient.
z x z z z
dx h(y) dy = h(y) dy dx = (z y) h(y) dy .
2 Relation to integration by parts a a a y a

This result can be seen to be an example of the formula

for integration by parts, as stated below:[4]
Consider the iterated integral

z z
f (x)g (x) dx = [f (x)g(x)]a f (x)g(x) dx
z x z
h(y) dy dx a a
a a
Substitute:
which we will write using the prex notation commonly
seen in physics:
x
f (x) = h(y) dy and g (x) = 1 .
z x a
dx h(y) dy Which gives the result.
a a

1
2 6 REFERENCES AND NOTES

3 Principal-value integrals regarded as functions of the corresponding parameter be,

respectively, continuous for c y < , a x < . Then
For application to principal-value integrals, see Whittaker if at least one of the iterated integrals
( ) ( )
and Watson,[5] Gakhov,[6] Lu,[7] or Zwillinger.[8] See also dy a dx f (x, y) and a dx c dy f (x, y)
c
the discussion of the Poincar-Bertrand transformation in
Obolashvili.[9] An example where the order of integration
converges, the other integral also converges and their val-
cannot be exchanged is given by Kanwal:[10]
ues coincide.

Theorem II Let f(x, y) be continuous for a x < , c

1 d1 d 1 y < , and let the integrals
g( ) = g(t) ,
(2i)2 L 1 t L 1 4 J(y) := a dx f (x, y) and J (x) = c dy f (x, y)
while:
be respectively, uniformly convergent on every nite in-
( )
terval c y < C and on every nite interval a x < A.
1 d1 Then if at least one of the iterated integrals
g( ) d =0.
(2i)2 L L (1 t) ( 1 )
( )
dy dx |f (x, y)| and

c ( a )
The second form is evaluated using a partial fraction ex- a dx c dy |f (x, y)|
pansion and an evaluation using the SokhotskiPlemelj
formula:[11]
converges, the iterated integrals
( ) ( )
dy a dx f (x, y) and a dx c dy f (x, y)
c
d1 d1
= =i.
L 1 t L 1 t also converge and their values are equal.

The notation L indicates a Cauchy principal value. See The most important theorem for the applications is
Kanwal.[10] quoted from Protter and Morrey:[14]
Theorem Suppose F is a region given by F =
{(x, y) : a x b, p(x) y q(x)} where p and q
4 Basic theorems are continuous and p(x) q(x) for a x b. Suppose
that f(x, y) is continuous on F. Then
A good discussion of the basis for reversing the order of f (x, y)dA = b q(x) f (x, y) dy dx .
F a p(x)
integration is found in the book Fourier Analysis by T.W.
[12]
Krner. He introduces his discussion with an example The corresponding result holds if the closed
where interchange of integration leads to two dierent region F has the representation F =
answers because the conditions of Theorem II below are {(x, y) : c y d, r(y) x s(y)} where r(y)
not satised. Here is the example: s(y) for c y d. In such a case,

[ ] d s(y)
x2 y 2 y 1
2 dy = x2 + y 2 = [x 1] . f (x, y)dA = f (x, y) dx dy .
1
2 2
(x + y ) 1 1 + x2 F c r(y)

( ) In other words, both iterated integrals, when computable,

x2 y 2 are equal to the double integral and therefore equal to
2 dy dx = . each other.
1 1 (x2 + y 2 ) 4
( )

x2 y 2
dx dy = .
1 1 (x2 + y2 )
2 4 5 See also
Two basic theorems governing admissibility of the inter- Fubinis theorem
change are quoted below from Chaudhry and Zubair:[13]
Theorem I Let f(x, y) be a continuous function of con-
stant sign dened for a x < , c y < , and let the 6 References and notes
integrals

J(y) := a dx f (x, y) and J (x) = c dy f (x, y) [1] Sen Dineen (2001). Multivariate Calculus and Geometry.
Springer. p. 162. ISBN 1-85233-472-X.
 3

[2] Richard Courant & Fritz John (2000). Introduction to Cal- Duane Nykamps University of Minnesota website
culus and Analysis: Vol. II/1, II/2. Classics in mathemat-
ics. Springer. p. 897. ISBN 3-540-66569-2. A general introduction.

[3] Department of Mathematics, Oregon State University.

Double Integrals.

[4] The prime notation " " denotes a derivative.

[5] Edmund Taylor Whittaker & George Neville Watson

(1927). A course of modern analysis : an introduction
to the general theory of innite processes and of analytic
functions, with an account of the principal transcendental
functions (4th ed., repr ed.). Cambridge University Press.
p. 4.51, p. 75. ISBN 0-521-58807-3.

[6] F. D. Gakhov (1990). Boundary Value Problems. Courier

Dover Publications. p. 46. ISBN 0-486-66275-6.

[7] Jian-Ke Lu (1993). Boundary Value Problems for Ana-

lytic Functions. Singapore: World Scientic. p. 44. ISBN
981-02-1020-5.

[8] Daniel Zwillinger (1992). Handbook of integration. AK

Peters Ltd. p. 61. ISBN 0-86720-293-9.

[9] Elena Irodionovna Obolashvili (2003). Higher order par-

tial dierential equations in Cliord analysis: eective so-
lutions to problems. Birkhuser. p. 101. ISBN 0-8176-
4286-2.

[10] Ram P. Kanwal (1996). Linear Integral Equations: theory

and technique (2nd ed.). Boston: Birkhuser. p. 194.
ISBN 0-8176-3940-3.

[11] For a discussion of the Sokhotski-Plemelj formula see, for

example, Joseph A. Cima, Alec L. Matheson & William
T. Ross (2006). The Cauchy Transform. American Math-
ematical Society. p. 56. ISBN 0-8218-3871-7. or
Rainer Kress (1999). Linear integral equations (2nd ed.).
Springer. p. Theorem 7.6, p. 101. ISBN 0-387-98700-2.

[12] Thomas William Krner (1988). Fourier Analysis. Cam-

bridge University Press. p. Chapters 47 & 48. ISBN 0-
521-38991-7.

[13] M. Aslam Chaudhry & Syed M. Zubair (2001). On a

Class of Incomplete Gamma Functions with Applications.
CRC Press. p. Appendix C. ISBN 1-58488-143-7.

[14] Murray H. Protter & Charles B. Morrey, Jr. (1985).

Intermediate Calculus. Springer. p. 307. ISBN 0-387-
96058-9.

7 External links
Pauls Online Math Notes: Calculus III

Good 3D images showing the computation of Dou-

ble Integrals using iterated integrals, the Depart-
ment of Mathematics at Oregon State University.

Ron Miechs UCLA Calculus Problems More com-

plex examples of changing the order of integration
(see Problems 33, 35, 37, 39, 41 & 43)
4 8 TEXT AND IMAGE SOURCES, CONTRIBUTORS, AND LICENSES

8 Text and image sources, contributors, and licenses

8.1 Text
Order of integration (calculus) Source: https://en.wikipedia.org/wiki/Order_of_integration_(calculus)?oldid=753575524 Contributors:
Michael Hardy, Giftlite, Sietse, Bender235, 3mta3, Kusma, Oleg Alexandrov, Rjwilmsi, Sodin, Petter Strandmark, SmackBot, Foxjwill,
Iridescent, MOBle, Alaibot, OrenBochman, Quietbritishjim, Plastikspork, Brews ohare, Franklin.vp, Addbot, More random musing, TStein,
Jasper Deng, Luckas-bot, Yobot, ShellyT123, Distantcube, Plusdrama33, SporkBot, IznoRepeat, Frietjes, Helpful Pixie Bot, BattyBot,
Dexbot, CsDix, CheChe and Anonymous: 10

8.2 Images
File:Integration_Order.svg Source: https://upload.wikimedia.org/wikipedia/commons/6/67/Integration_Order.svg License: CC BY-SA
4.0 Contributors: This le was derived from Integration Order.PNG: <a href='//commons.wikimedia.org/wiki/File:Integration_Order.
PNG' class='image'><img alt='Integration Order.PNG' src='https://upload.wikimedia.org/wikipedia/commons/thumb/4/4f/Integration_
Order.PNG/50px-Integration_Order.PNG' width='50' height='24' srcset='https://upload.wikimedia.org/wikipedia/commons/thumb/4/4f/
Integration_Order.PNG/75px-Integration_Order.PNG 1.5x, https://upload.wikimedia.org/wikipedia/commons/thumb/4/4f/Integration_
Order.PNG/100px-Integration_Order.PNG 2x' data-le-width='921' data-le-height='444' /></a>
Original artist: CheCheDaWa

8.3 Content license

Creative Commons Attribution-Share Alike 3.0