Fact-checked by Grok 2 weeks ago

Principal value

In mathematics, particularly in , the principal value of a is the specific value obtained by restricting the function to its , which selects a representative from the set of possible values, often using the of the in the interval (-\pi, \pi]. This approach resolves the inherent ambiguity of functions like the logarithm or roots, enabling single-valued definitions on domains such as the excluding a , typically along the negative real axis. A prominent example is the principal value of the , denoted \operatorname{Log}(z), defined for z \neq 0 as \operatorname{Log}(z) = \ln |z| + i \operatorname{[Arg](/page/ARG)}(z), where \operatorname{[Arg](/page/ARG)}(z) is the principal argument satisfying -\pi < \operatorname{[Arg](/page/ARG)}(z) \leq \pi. This principal branch maps the punctured complex plane to a horizontal strip in the complex plane, ensuring continuity except across the branch cut, and serves as the inverse of the exponential function within that strip. Similarly, for the nth root function z^{1/n}, the principal value is |z|^{1/n} e^{i \operatorname{[Arg](/page/ARG)}(z)/n}, with the argument restricted to (-\pi/n, \pi/n], providing a continuous function on \mathbb{C} \setminus (-\infty, 0]. Beyond multivalued functions, the term principal value also refers to the in the theory of improper integrals, which assigns a finite value to integrals that diverge under standard limits by symmetrically excluding singularities. For a function f continuous except at x = a, the is defined as \operatorname{PV} \int_{-\infty}^{\infty} f(x) \, dx = \lim_{\epsilon \to 0^+} \left( \int_{-\infty}^{a - \epsilon} f(x) \, dx + \int_{a + \epsilon}^{\infty} f(x) \, dx \right), provided the limit exists. This construction converges even when the ordinary integral does not, as in the case of \int_{-\infty}^{\infty} \frac{1}{x} \, dx, where the principal value is 0 due to cancellation of symmetric contributions. It extends to multiple singularities via symmetric exclusions around each and plays a key role in residue theorem applications, distribution theory, and Fourier transforms. These principal value concepts are foundational in advanced mathematics, facilitating the analysis of singularities, branch points, and divergent expressions while preserving essential symmetries and analytic properties.

Conceptual Foundations

Multi-valued Functions in Complex Analysis

In complex analysis, certain functions such as the logarithm and square root are inherently multi-valued when defined over the complex plane, meaning that a single input z \neq 0 can yield multiple distinct outputs due to the periodic and topological structure of the complex domain. This multi-valuedness arises because the complex plane can be represented in polar form as z = re^{i\theta}, where \theta is the argument, which is only defined up to multiples of $2\pi, leading to an infinite family of equivalent representations for each z. For instance, the complex logarithm, defined inversely to the exponential function via \exp(w) = z, satisfies \log(z) = \ln|z| + i \arg(z), but since \arg(z) differs by $2\pi k for any integer k, the logarithm takes infinitely many values \ln|z| + i (\arg(z) + 2\pi k). A key mechanism revealing this multi-valued nature is analytic continuation, which extends the domain of an analytic function beyond its initial definition while preserving holomorphicity. When analytically continuing a function like the logarithm around a closed path encircling the origin (a branch point at z=0), the value shifts by $2\pi i upon completing one full loop, as the argument accumulates an extra $2\pi. Repeated encirclements yield further increments of $2\pi i k, demonstrating how the topology of the punctured complex plane (\mathbb{C} \setminus \{0\})—which is not simply connected—forces the function to "wind" and produce distinct values, preventing a single-valued analytic extension over the entire plane without additional structure. This understanding of multi-valued functions traces back to Bernhard Riemann's foundational work in the 1850s, particularly his 1851 dissertation "Grundlagen für eine allgemeine Theorie der Functionen einer veränderlichen complexen Grösse," where he introduced concepts like Riemann surfaces to geometrically resolve the multi-valuedness of such functions using Cauchy-Riemann equations and integral representations. Riemann's 1857 paper on Abelian functions further advanced the theory by addressing inverses and periodicity in multi-valued contexts, influencing the modern treatment of these phenomena. A concrete illustration is the square root function, which is two-valued in the complex plane, assigning to each z \neq 0 the pair of solutions w such that w^2 = z. For example, at z = -1, the values are \pm i, corresponding to the two possible branches: one yielding i and the other -i, with the choice depending on the path taken to approach the point, underscoring the function's dependence on the complex plane's topology.

The Role of Branches

In complex analysis, a branch of a multi-valued function is defined as a single-valued analytic function that selects one value from the set of possible values at each point in a suitable domain, ensuring continuity and analyticity except possibly on the boundary, typically by excluding branch cuts from the complex plane. This construction addresses the inherent ambiguity of multi-valued functions, such as those arising from inverse operations like roots or logarithms, by restricting the domain to avoid discontinuities at branch points. Riemann surfaces offer a geometric framework for understanding branches globally, representing the multi-valued function as a single-valued one on a multi-sheeted covering of the complex plane. For the logarithmic function, the associated Riemann surface takes the form of an infinite helix, where each sheet corresponds to a successive increment of $2\pi i in the imaginary part, allowing seamless continuation around the branch point at the origin without jumps in value. This structure illustrates how branches can be extended analytically across the entire surface, providing a unified view of the function's behavior. Branches of a given multi-valued function are not unique; varying the choice of branch cut or the selection of values on each sheet produces distinct analytic functions, though they coincide in overlapping regions of their domains. For instance, the two branches of the differ by a multiplicative factor of -1 everywhere in the complex plane excluding the branch point. For the , with the branch cut along the negative real axis, the limiting values from the upper and lower sides of the cut are negatives of each other, illustrating the discontinuity while maintaining analyticity within the domain of each branch. This non-uniqueness underscores the role of branches in enabling local analytic continuation, bridging the gap between multi-valued expressions and practical single-valued computations in .

Defining the Principal Value

General Definition of Principal Branch

In complex analysis, the principal branch of a multi-valued analytic function is defined as a particular single-valued, continuous, and analytic branch selected from the family of possible branches to serve as a canonical representative. This choice resolves the inherent ambiguity arising from the periodic nature of multi-valued functions, such as those involving roots or logarithms, by restricting the function to a specific "slice" on the Riemann surface. The principal branch is typically identified by confining the argument of the complex variable to a principal range, most commonly the interval (-\pi, \pi] for functions like the . For a general multi-valued function expressible as f(z) = g(\ln |z| + i \theta), where g is a suitable analytic function and \theta = \arg z, the principal value is obtained by taking \theta within this principal interval, ensuring the result aligns with the standard real-valued extension. This restriction yields a well-defined function on the slit complex plane, excluding the branch cut. Selection of the principal branch adheres to specific criteria: it must be continuous with the corresponding real-valued function along the positive real axis, where the imaginary part vanishes, and it should maximize the domain of analyticity, typically the entire complex plane minus a ray from the origin along the branch cut (often the non-positive real axis). These criteria guarantee that the principal branch is holomorphic in this maximal simply connected domain and matches intuitive real-line behavior, such as \log x = \ln x for x > 0. Standardization of the principal branch is crucial for achieving uniqueness in both theoretical developments and practical computations within . Without a conventional , multi-valued functions would lack a consistent reference, complicating evaluations in series expansions, integrals, and numerical algorithms. Moreover, it underpins key results like the argument principle, which depends on a fixed to accurately determine the and thus the number of of analytic functions inside contours.

Conventions for Branch Cuts

Branch cuts are curves in the complex plane, often rays emanating from branch points or singularities, across which a multi-valued analytic function is intentionally made discontinuous to render it single-valued in the cut plane. This discontinuity enforces a specific branch of the function by preventing closed paths that encircle branch points, which would otherwise lead to multiple values. For principal branches of functions like the logarithm and non-integer powers, the standard convention places the branch cut along the negative real axis, ensuring that the principal argument satisfies \arg(z) \in (-\pi, \pi]. This choice aligns the principal branch with the real logarithm on the positive real axis and is widely adopted in mathematical software and texts for consistency. The negative real axis cut offers the advantage of around the real axis, preserving real values for positive real inputs and simplifying extensions from real to ; however, it introduces a discontinuity that can affect the analyticity in regions crossing the negative reals, potentially requiring careful path deformation in integrals. In comparison, a positive real axis cut, yielding \arg(z) \in [0, 2\pi), provides an asymmetric range that may better suit applications emphasizing the upper half-plane but disrupts continuity on the positive reals, making it less common for principal branches. For a general algebraic function such as w = z^{1/2} (z-1)^{1/3}, branch points occur at z=0 and z=1, with an additional point at infinity; cuts are typically placed as a ray from z=0 to infinity along the positive real axis and a segment from z=1 to infinity in another direction, ensuring no branch point is enclosed by any closed contour in the cut plane to maintain single-valuedness.

Applications to Elementary Functions

Principal Logarithm

The principal branch of the , denoted \Log z, is defined for z \in \mathbb{C} \setminus \{0\} by the formula \Log z = \ln |z| + i \Arg z, where \Arg z is the principal argument of z, restricted to the interval (-\pi, \pi]. This definition selects a single-valued representative from the multi-valued \log z = \ln |z| + i \arg z + 2\pi i k for k \in \mathbb{Z}, ensuring continuity and analyticity in the excluding the branch cut along the non-positive real axis. The function \Log z is analytic (holomorphic) on \mathbb{C} minus the non-positive real axis, where it serves as the inverse of the restricted to the horizontal strip \{w \in \mathbb{C} : -\pi < \Im w \leq \pi\}. Its derivative is \frac{d}{dz} \Log z = \frac{1}{z}, which holds throughout the domain of analyticity. The multiplicative property is \Log(zw) = \Log z + \Log w + 2\pi i k, where k = 0 or $1 (or occasionally -1) depending on whether the sum of the principal arguments \Arg z + \Arg w lies within (-\pi, \pi] or requires adjustment by \pm 2\pi to return to that interval; the property holds exactly without the correction term when z and w are chosen such that zw does not cross the branch cut. For positive real numbers x > 0, \Arg x = 0, so \Log x = \ln x, aligning with the real . Examples illustrate this branch: \Log(-1) = i\pi since |-1| = 1 and \Arg(-1) = \pi, and \Log i = i \pi / 2 since |i| = 1 and \Arg i = \pi / 2. However, \Log z exhibits a discontinuity across the branch cut on the non-positive real axis: as z approaches a negative real number from above (positive imaginary part), \Arg z \to \pi, while from below, \Arg z \to -\pi, resulting in a jump of $2\pi i in the value of \Log z. This discontinuity underscores the principal branch's role in providing a consistent analytic continuation of the logarithm while respecting the topology of the punctured complex plane.

Principal Square Root

The principal square root of a complex number z \neq 0, denoted \sqrt{z}, is defined as \sqrt{z} = \sqrt{|z|} \, e^{i \Arg(z)/2}, where \Arg(z) is the principal argument of z with range (-\pi, \pi]. This choice ensures that the principal value has a non-negative real part, as the argument of \sqrt{z} lies in (-\pi/2, \pi/2]. The function can also be expressed in Cartesian form as \sqrt{z} = \sqrt{|z|} \left( \cos(\Arg(z)/2) + i \sin(\Arg(z)/2) \right). For z = 0, the principal square root is defined as 0. The has branch points at z = 0 and z = \infty, where it fails to be single-valued under continuous variation around these points. To define the principal branch, a branch cut is introduced along the negative real axis, making the function discontinuous across this from 0 to -\infty. This principal branch is analytic everywhere in the except on the branch cut and at the branch points. Its , where defined, is \frac{d}{dz} \sqrt{z} = \frac{1}{2 \sqrt{z}}. Key properties of square root include the identity \sqrt{z^2} = |z| for all z \in \mathbb{C}, rather than z itself, due to the choice that prioritizes the non-negative real part. This reflects the multi-valued nature resolved by the principal . The principal square root can be derived from the principal logarithm via , as \sqrt{z} = \exp\left( \frac{1}{2} \Log z \right). For example, \sqrt{-1} = i, since -1 = e^{i\pi} and \sqrt{-1} = e^{i\pi/2} = i. Similarly, \sqrt{-4} = 2i, as -4 = 4 e^{i\pi} yields \sqrt{-4} = 2 e^{i\pi/2} = 2i.

Principal Values of Inverse Trigonometric Functions

The principal values of the inverse trigonometric functions arcsin(z), arccos(z), and arctan(z) are defined to provide single-valued branches in the complex plane, ensuring continuity and matching the conventional real-valued ranges for arguments in the appropriate intervals. For real z ∈ [-1, 1], the principal value of arcsin(z) lies in [-π/2, π/2], arccos(z) in [0, π], and arctan(z) in (-π/2, π/2). These ranges are chosen to make the functions strictly increasing and to cover the full period of the sine and cosine functions without redundancy. In the complex domain, the principal branch of arcsin(z) is given by \arcsin z = -i \log \left( i z + \sqrt{1 - z^2} \right), where log denotes logarithm and the square root is the principal branch. This definition uses the principal square root, which has nonnegative real part, to ensure the argument of the logarithm lies in the principal strip. The branch cut for arcsin(z) is along the real axis from -∞ to -1 and from 1 to ∞. Similarly, the principal branch of has range [0, π] for real z ∈ [-1, 1] and is defined as \arccos z = -i \log \left( z + i \sqrt{1 - z^2} \right) or equivalently, \arccos z = \pi/2 - \arcsin z, with the same branch cuts along (-∞, -1] ∪ [1, ∞). This relation holds on the principal branches, preserving the complementary nature of sine and cosine. For arctan(z), the principal value lies in (-π/2, π/2) for real z, and in the it is \arctan z = \frac{i}{2} \log \left( \frac{i + z}{i - z} \right) = \frac{i}{2} \log \left( \frac{1 - i z}{1 + i z} \right), with branch cuts along the imaginary from -i∞ to -i and from i to i∞. These cuts avoid the poles of the function at odd multiples of iπ/2. As an illustrative example, the principal value of arcsin(2) is \pi/2 - i \ln(2 + \sqrt{3}) \approx 1.5708 - 1.3170 i, which lies outside interval but respects the structure.

Principal Values of

The principal values of in the are defined using the principal branch of the logarithm, ensuring single-valued analytic continuations except across specified branch cuts. These functions are multi-valued due to the inherent branching of the logarithm and square roots involved in their expressions, but the principal branch selects a specific range that aligns with the real-valued outputs for appropriate real inputs, distinguishing them by their unbounded nature in the real direction or infinite strips parallel to the real axis. For the inverse hyperbolic sine, the principal value is given by \operatorname{arcsinh} z = \log\left(z + \sqrt{z^2 + 1}\right), where \log denotes the principal logarithm with argument in (-\pi, \pi]. This yields real values for real z, spanning the range (-\infty, \infty), and the function is analytic in the except for branch cuts along the imaginary axis from -i\infty to -i and from i to i\infty. The branch points are at z = \pm i. The principal value of the inverse hyperbolic cosine is \operatorname{arccosh} z = \log\left(z + \sqrt{z^2 - 1}\right), with the square root and logarithm taken as principal branches; for \operatorname{Re} z > 0, the positive root is selected to ensure the principal value is real and non-negative for real z \geq 1, spanning [0, \infty). The branch cut lies along the real axis from -\infty to $1, with branch points at z = \pm 1. In the complex plane, the principal range forms a strip [0, \infty) \times [-\pi, \pi]$. For the inverse hyperbolic tangent, the principal value is \operatorname{arctanh} z = \frac{1}{2} \log\left(\frac{1 + z}{1 - z}\right), again using the principal logarithm. This produces real values spanning (-\infty, \infty) for real z with |z| < 1, and the principal branch in the complex plane lies within the infinite strip where the imaginary part is in (-\pi/2, \pi/2). Branch cuts run along the real axis from -\infty to -1 and from $1to\infty, with branch points at z = \pm 1$. An illustrative example is \operatorname{arcsinh}(i) = i \cdot \pi/2, which follows from the relation to the sine on the of the and highlights the imaginary-valued output within the range. Branch cuts for these functions are typically segments on the real or , tailored to avoid the paths where the real-valued definitions hold.

The Argument Function

Definition and Principal Range

In , the argument of a non-zero z, denoted \arg(z), is defined as the of the : \arg(z) = \Im(\Log(z)), where \Log(z) is the principal branch of the . This represents the angle, measured counterclockwise from the positive real to the ray from the origin through z in the . The principal value of the argument, denoted \Arg(z), is the unique value of \arg(z) lying in the interval (-\pi, \pi]. This convention introduces a discontinuity along the non-positive real axis, where \Arg(z) approaches \pi from the upper half-plane and -\pi from the lower half-plane as z approaches the cut. A key property of the argument is its additivity under multiplication: \arg(z w) = \arg(z) + \arg(w) modulo $2\pi, meaning \Arg(z w) = \Arg(z) + \Arg(w) + 2\pi k for some k chosen to ensure the result lies in (-\pi, \pi]. The principal argument function \Arg(z) is single-valued but discontinuous along the branch cut on the non-positive real axis, hence nowhere analytic. It is continuous everywhere else in the . The full argument is multi-valued, with branches differing by integer multiples of $2\pi. For example, \Arg(-1) = \pi, corresponding to the angle at the negative real axis endpoint of the principal range. Similarly, for z = e^{i 3\pi/2} = -i, \Arg(z) = -\pi/2 rather than $3\pi/2, as the latter exceeds the principal interval and is reduced modulo $2\pi.

Relation to Principal Logarithm

The principal logarithm of a complex number z \neq 0, denoted \Log(z), is defined by the formula \Log(z) = \ln |z| + i \Arg(z), where \Arg(z) is the principal value of the argument with range (-\pi, \pi]. This relation establishes that the principal argument uniquely determines the principal branch of the logarithm, rendering \Log(z) analytic in the complex plane excluding the branch cut along the non-positive real axis. This foundational connection extends to other multi-valued functions, particularly powers of numbers. For a exponent a, the principal value of z^a is given by z^a = \exp(a \Log(z)), which inherits the branch cut and discontinuity from \Log(z) and \Arg(z). Consequently, z^a is single-valued and continuous on the principal branch domain, but encircling the introduces additional values differing by factors of \exp(2\pi i n a) for n. Crossing the branch cut, such as the negative real axis, causes \Arg(z) to exhibit a jump discontinuity of $2\pi, which in turn shifts \Log(z) by $2\pi i. This discontinuity propagates to dependent functions like powers, disrupting their across the cut and necessitating careful domain restrictions for analyticity. In applications to , paths are constructed to avoid the branch cut and maintain the principal values of \Log(z) and \Arg(z). A representative example is the keyhole contour, which encircles the while indenting around the positive real axis cut; this setup enables applications for integrals involving principal branches, such as evaluating \int_0^\infty \frac{x^{\alpha-1}}{1+x} \, dx for $0 < \alpha < 1 by considering the on the cut-avoiding domain.