On the slit plane , choose
so that as . This branch cut gives, for , the boundary values
Integrate over the boundary of a large disk with a thin slit removed. The upper bank is traversed from to and the lower bank from to , so together they contribute
At infinity the Laurent series is
and hence the counterclockwise large-circle integral tends to . By the Cauchy integral theorem, the large-circle and slit-boundary contributions sum to zero. Therefore
which yields
Jordan lemma states that if and is analytic in the upper half-plane outside a fixed circle, with
then on the upper semicircle ,
Indeed, . Using symmetry and on gives
The residue of at an isolated singularity is the coefficient of in its Laurent series, equivalently
If , the Taylor expansion of the analytic function shows that
For the real integral, apply contour integration in the upper half-plane to
Jordan's lemma removes the semicircular arc. The only enclosed singularity is the double pole at , whose residue is
The residue theorem therefore gives
Its real part vanishes because is an odd function, while its imaginary part is the requested integral. Hence
A Laurent series about is an expansion
that converges to on an annulus .
The partial fraction decomposition is
For , expand both denominators by the geometric series:
This is actually a Taylor series at zero, since the apparent puncture at zero contains no singularity.
For , the term still expands in nonnegative powers, while the quadratic term must be expanded in negative powers:
Hence
An analytic branch of a square root determined by the branch is
If and are two such branches, their ratio is analytic and satisfies . Since a domain is connected and takes values in the discrete set , is constant. Thus throughout or throughout .
For , the principal square root is
On , one may instead take
The respective removed half-axes are their branch cuts.
On , define
It is analytic there, its square is , and
so it is the required branch. Substitution gives, for ,
The binomial series yields
hence the first three terms of the Laurent series are
Since
its residue at zero is . Under the change of variable , the two orientation reversals cancel, and the residue theorem gives
For , expand directly in inverse powers:
Therefore the outer Laurent series is
The poles of inside are . Near , write . Its Laurent series begins
so the residue is . The sum of the three residues is
The positively oriented residue theorem now gives
For each nonzero period lattice point , the binomial series gives
Pairing with cancels every odd power. Hence the Laurent series of the Weierstrass elliptic function is
which is the stated coefficient formula.
The Laurent theorem says that if is analytic on an annulus
then it has a unique Laurent series
converging locally uniformly on that annulus, where
for any positively oriented circle in the annulus around .
An isolated singularity at is a point at which is not analytic although it is analytic on some punctured neighbourhood. It is removable when every with vanishes; it is a pole of order when and for ; and it is essential when infinitely many negative-index coefficients are nonzero.
For ,
For ,
The coefficients are unique after the annulus is fixed; these expansions differ because they represent the function on different annuli. At zero the first expansion has principal part , so zero is a simple pole with residue .
For
a finite point is an ordinary point exactly when and are holomorphic there. It is a regular singular point exactly when
are holomorphic at .
Put and . Direct differentiation gives
so the transformed equation is
Consequently is ordinary precisely when
are holomorphic at . It is a regular singular point at infinity precisely when and are holomorphic functions of near infinity.
If zero and infinity are regular singular and every nonzero finite point is ordinary, the Laurent series of and can contain only the terms compatible with both endpoint bounds. Hence
for constants . The equation is a Cauchy-Euler differential equation. Its indicial equation is
For distinct roots , the general solution on a domain with a chosen logarithm branch is
For a repeated root , the general solution of an Euler-Cauchy equation is
Finally require infinity to be ordinary. In the transformed equation its coefficients become
Both are holomorphic at zero exactly when and . Thus the further restriction is
The classification of isolated singularities can be stated through the Laurent expansion of about :
  • The singularity is removable when extends holomorphically to , equivalently when every negative-power Laurent coefficient is zero.
  • It is a pole of order when extends holomorphically to a function nonzero at . Equivalently, the Laurent series starts with a nonzero term and has no more negative power.
  • It is essential when it is neither removable nor a pole, equivalently when infinitely many negative-power coefficients are nonzero.
The poles inside are zero and for . At a nonzero pole,
and the residue at is its negative. At zero, the Laurent series contains only even powers because the function is even, so its residue is zero. All residues cancel and therefore
Put . The Laurent series of the logarithmic derivative at the simple pole of has the form
where in fact . Hence
The coefficient of in their product, and therefore the residue, is
with the constant ; equivalently, .
Insert into the exponential series . For every fixed , this gives the expansion with absolute convergence
This is a Laurent series in inverse powers of , as requested. It is not a Maclaurin series about zero: that series has every coefficient zero and cannot equal on a punctured neighborhood of zero. There is no contradiction, since the inverse-power expansion does not apply at .
Poles and residues. An isolated singularity is a pole of order if its Laurent series has finitely many negative powers, the lowest being a nonzero multiple of . The residue is the coefficient of .
If and , factor , where is holomorphic near and . For ,
The last quotient is holomorphic and nonzero at , proving a simple pole with
The given function. For noninteger , specify a branch of a multivalued function before using . Choose with , a cut along the negative imaginary axis. This leaves all three roots of off the cut. They are , and , all simple. The numerator is nonzero at these points. Since the denominator has derivative , the residues are
For noninteger , zero is a branch point, rather than an isolated pole. This remains true when and the numerator diverges there. For nonnegative integer it is a regular point. Other logarithm branches change the phases of the residues consistently.
Integral and convergence. Near zero the integrand behaves as , requiring . At infinity it behaves as , requiring . Therefore the exact convergence range is
Use a counterclockwise sector contour with rays at angles and , inner radius and outer radius . The same complex logarithm branch is valid throughout this sector. Only lies inside. The small circular arc contributes and the large one contributes , so both vanish precisely in the stated range.
The outward positive-real ray contributes . On the returning ray set : the denominator is still , while , with reversed limits. The residue theorem gives
Put . The right-hand phase is and . Hence
The sine is positive in this range, consistent with a positive real integral. This is a special case of the Mellin integral of a reciprocal binomial.
Definitions. In a punctured neighborhood of , let the Laurent series be . The point is a pole of order if and every coefficient with index less than vanishes. Its residue is . A simple pole has .
Poles for . Denote the bracket in the denominator by . Factoring gives
Put , so . The finite poles are , all simple; the numerator is nonzero at each. At zero,
At a nonzero simple root of , the residue formula gives
For , ; for , . Substitution yields
The two zeros of a function at do not cause additional poles: there and the numerator vanishes.
The case . Simplify before using any formulas that divide by :
Thus its only finite pole is
As , the two inner poles merge with zero while the outer two escape to infinity. If one also considers the Riemann sphere, at has a simple pole at infinity because ; the requested unit-circle calculation only uses finite poles.
Unit-circle integral. Set . Then
At the two angles where the tangent is undefined, the real integrand has continuous limiting value zero. Thus the substitution is valid for the improper integral, or after that continuous extension. No pole lies on for any allowed .
For , the sum of the enclosed residues is
The residue theorem therefore gives
At the enclosed residue is , giving , exactly the same formula. The direct simplification provides an independent check of that endpoint.
Compute the complex derivative along real increments to get , and along imaginary increments to get . Equality gives the Cauchy-Riemann equations, .
For the specified real part, observe that . Thus the most general holomorphic function with this real part on the connected plane is
The difference of two such holomorphic functions has zero real part, so the Cauchy-Riemann equations make its imaginary part constant.
Next, . Its real and imaginary parts satisfy and . Their partial derivatives are continuous everywhere, proving that the complex hyperbolic cosine is entire.
For the Laurent series, set . Since ,
The regular part at a simple pole therefore gives
A pole of order at has extending holomorphically to a nonzero value there. Its residue is the coefficient of in the Laurent series. The Cauchy integral theorem gives for a closed contour in a simply connected domain of holomorphicity. The residue theorem replaces zero by times the enclosed residues for a positively oriented simple contour avoiding finitely many poles.
For the first integral, use the upper-half-plane semicircle, with a small clockwise upper semicircle excluding the origin. Near zero,
There are no poles inside the indented contour. The large arc vanishes because and the integrand there is ; the small arc tends to . Therefore the real-axis complex principal value is . Taking real parts gives the ordinary convergent cosine-difference inverse-square integral
The real numerator cancels to second order at zero, so this final real integral itself needs no principal-value convention.
For , the simple poles are , . Differentiating the denominator gives
Integrate around the rectangle with vertices . It contains only . On the top side , and reversed traversal therefore contributes times the bottom integral. The vertical sides vanish for . The residue theorem gives , so the exponential moment of the hyperbolic secant is
Figure 1.
Indented semicircle and height-one rectangle for the two real integrals
. Original contour diagrams. The left contour excludes the origin with a clockwise indentation; the right encloses only the pole . Other hyperbolic-secant poles are shown outside the rectangle. Details.
Put . Since and , the local Laurent series is
Thus all singular terms are
There is a third-order pole, with no second-order term.
For a simple pole, the residue is , the coefficient of its Laurent series. On an anticlockwise circle , integration of each Laurent term is zero except , whose integral is . This proves the stated single-pole residue theorem.
For , the only pole inside the rectangle of height is , with residue . The residue theorem therefore gives . The upper edge is traversed from right to left and satisfies , so the two horizontal contributions sum to .
On the right edge the modulus is bounded by , and on the left by . Multiplying by the fixed edge length shows that both contributions tend to zero because . Thus
Using yields
This rectangular contour for an exponential beta integral also gives the ordinary beta integral after the substitution .