when the limit exists (Henrici 1988, p. 261; Whittaker and Watson 1990, p. 117; Bronshtein and Semendyayev 1997, p. 283). Similarly,
the Cauchy principal value of a doubly infiniteintegral
of a function is defined by
when the limit exists. The Cauchy principal value is also known as the principal value integral (Henrici 1988, p. 261). For a simple
pole in the interior of the integration interval,
it agrees with the corresponding Hadamard finite part.
In general, however, the finite part and the Cauchy
principal value are distinct notions (Galapon 2016).
The Cauchy principal value of an integral having no nonsimple poles can be computed in the Wolfram
Language using Integrate[f,
x,
a, b,
PrincipalValue -> True]. Cauchy principal values of functions
with possibly nonsimple poles can be computed numerically
using the "PrincipalValue" method in NIntegrate.
Cauchy principal values are important in the theory of generalized functions, where they allow extension of results to .
Cauchy principal values are sometimes simply known as "principal values" (e.g., Vladimirov 1971, p. 75) even though they are not related to the principal
value of complex analysis.
The most common designation for the Cauchy principal values seems to be (Henrici 1988, pp. 259-262; Gradshteyn and
Ryzhik 2000, p. 523). Sometimes, no explicit designation is used (Harris and
Stocker 1998, p. 552; Gradshteyn and Ryzhik 2000, p. 248). Other notations
include
(Arfken 1985, p. 403), P.V. (Apelblat 1983, p. viii), (Morse and Feshbach 1953, p. 368; most Russian authors),
(Vladimirov 1971), (Bronshtein and Semendyayev 1997, p. 282), and V.P.
(Brychkov 1992, p. 7). For integrals with finite
integration limits, the Cauchy principal value is sometimes denoted (Zwillinger 1995, p. 346).