Choose the trigonometric substitution with . The cosine is nonnegative there, so and . The double-angle formula then gives
Since and ,
The formula is an antiderivative on and has continuous endpoint values at .
The product in the integrand simplifies to below its upper endpoint. With the integration by substitution , the limits become and zero and . Therefore
Although the tangent function itself is undefined at the upper endpoint, the product has a zero limit there, so its continuous extension, or its improper integral, gives this same result.
To obtain directly from a Riemann sum, divide into equal pieces of width and use right endpoints . The sum is
The integrand is continuous, so the limit of these Riemann sums is its definite integral, namely , agreeing with the antiderivative calculation.
Use integration by substitution with , so and . The indefinite integral becomes
Thus, with an arbitrary constant of integration,
This real antiderivative applies for and extends continuously to the stated endpoint. Its left derivative there agrees with the limiting integrand, zero.
A suitable real antiderivative on the disconnected domain is
For , directly; for , and the chain rule again gives . The graph is even, crosses zero at , tends to on either side of zero, and tends to as . Independent additive constants may be chosen on the two disconnected half-lines; both are set to zero here.
Figure 1.
Real antiderivative branches for and , with their domains and limiting behavior
. Details.
Use integration by substitution with , so . Then
Differentiating the answer by the chain rule gives , confirming the antiderivative. The real constant of integration is arbitrary.
With held fixed, the inner antiderivative gives
The remaining integral is therefore
Write and . Two applications of integration by parts, first taking the sine and then the cosine as the differentiated factor, give
Thus and , in each case up to an arbitrary constant of integration. Keeping that freedom avoids identifying independently chosen antiderivatives too rigidly.
For , write and apply integration by parts, using the antiderivative for the second factor:
The boundary term is zero: at zero, , and at infinity Gaussian decay dominates every power. Consequently the Gaussian moment recurrence is
First suppose . Differentiation under the integral sign, justified on compact positive parameter intervals by an integrable bound, gives
The first integral is . For , use the partial fraction decomposition
The standard inverse tangent antiderivatives then give
The same value holds at by continuity, or by direct integration. Hence the proposed first-order linear differential equation is valid on the positive parameter range:
Divide by to recognise a quotient rule derivative:
To fix , the value alone is insufficient, since every displayed solution tends to zero there. Instead consider
This limit follows from the dominated convergence theorem: for and , the logarithm's absolute value is bounded by ; for , it is at most . After division by , these are integrable. The limiting improper integral is absolutely convergent, and the substitution rule changes it into its own negative. Its value is therefore zero, giving . Thus
The question calls a real parameter, but this last formula and the proposed differential equation do not hold on the entire real axis. The defining integral is odd in , because its logarithm depends on . Applying the positive-parameter result to therefore gives the correct extension
with . In particular, for , the right-hand side of the differential equation must be ; for every the unified expression is . The logarithmic quadratic parameter integral therefore provides the requested identity on its valid range and the corrected answer elsewhere.
On a nonconstant branch with , regard as a function of . The equation becomes the linear ordinary differential equation
An integrating factor is , giving
Taking an antiderivative yields the implicit family
On a regular branch this can also be written . The latter form loses the constant solutions , which must be restored separately.
The implicit relation is a first integral of an ordinary differential equation: along any differentiable solution of the original equation,
Conversely, a differentiable graph satisfying this relation solves the original equation wherever . The original equation is undefined when ; points at which also need separate attention before solving explicitly for . This is why dividing immediately by can lose valid solutions and obscure nonuniqueness.
The function graphs intersect when , so or . For , the exponential function satisfies , making the upper graph. The enclosed area is therefore
The antiderivative of follows by integration by parts.
At height , the circular sector has radius and angle . Its area is , so slicing by horizontal cross-sections gives the volume integral
Use the substitution rule . Then
An integration by parts gives the antiderivative , whose endpoint difference is . Therefore
On an interval with , division by gives the separable differential equation . For , take antiderivatives:
Renaming the constant yields the explicit nonzero family
For , the interval of definition must avoid zeros of the denominator. Division by excluded the additional solution , which must also be included.
The original undivided differential equation also makes sense at , where it requires . For continuation through a zero of the leading ODE coefficient, rewrite the nonzero family as
It extends differentiably with and , since . One may choose the integration constants independently on the two sides of zero and obtain a differentiable solution through zero. The zero solution also extends through zero, with derivative ; joining a nonzero branch on just one side to a zero branch on the other would fail differentiability.
For , the even function has , because is an odd function. Its constant coefficient is . The remaining Fourier coefficients are
Two applications of integration by parts give the antiderivative, with ,
Its endpoint difference is , so . Therefore
The periodic extension is continuous and piecewise smooth, so pointwise convergence of a piecewise smooth Fourier series gives equality, including at the endpoints.
The partial fraction decomposition is
Taking antiderivatives and using the positive endpoints gives
Hence .
Write . Apply integration by parts with and :
Using on the right gives . Rearranging proves
For , the analogous integration by parts uses and , giving . Hence
An arbitrary constant of integration absorbs the choices in the lower-order antiderivatives.
Starting from , successive use of gives
These expressions follow directly from the requested recurrence, rather than requiring a separate trigonometric expansion.
With , divide the separated equation on , to obtain
Thus and . Taking antiderivatives,
Choose and the signed branch . A nonconstant solution, smooth for every real , is
Indeed and , so . This verification includes the axes excluded temporarily during separation.