0
$\begingroup$

I have the following function (it is the incomplete elliptic integral of first kind) $$ F(b,g) = \int_{0}^{b} \frac{dx}{\sqrt{(1-x^2)(1-gx^2)}} $$ I would like to compute $$\frac{\partial F}{\partial g} \ ,\ \frac{\partial F}{\partial b} \ ,\ \frac{\partial^2 F}{\partial g^2} \ ,\ \frac{\partial^2 F}{\partial b^2} \ ,\ \frac{\partial^2 F}{\partial b\partial g}$$ so I defined

F[b_,g_]:= Integrate[1/Sqrt[(1 - x^2)*(1 - g*x^2)], {x, 0, b}]

and tried the command

D[F[b,g],g]

but Mathematica cannot compute it. Am I doing something wrong or is there a way to do it?

$\endgroup$
4
  • 1
    $\begingroup$ Why not use EllipticF instead of an integal? E.g., F[b_, g_] := EllipticF[ArcSin[b], g]. $\endgroup$
    – Carl Woll
    Oct 20 '20 at 16:26
  • $\begingroup$ An indirect way: try Integrate[f[x, g], {x, 0, b}] // D[#, g] & , then replace f with real func. $\endgroup$
    – wuyudi
    Oct 20 '20 at 16:31
  • $\begingroup$ @CarlWoll because I want to keep it as general as possible in order to modify it easily someday $\endgroup$ Oct 20 '20 at 16:36
  • $\begingroup$ Derivatives relative to "b" are trivial and why not swap the derivative relative to g and the integral? $\endgroup$ Oct 20 '20 at 16:46
2
$\begingroup$

I don't know if you want to be this restrictive, but you can avoid singularities that Mathematica evidently has trouble with by making assumptions.

F[b_,g_]:= Integrate[1/Sqrt[(1 - x^2)*(1 - g*x^2)], {x, 0, b}]

$Assumptions = 0 < b < 1 && 0 < g < 1

D[F[b, g], g]

$\frac{\sin \left(2 \sin ^{-1}(b)\right)}{4 (g-1) \sqrt{1-b^2 g}}-\frac{F\left(\left.\sin ^{-1}(b)\right|g\right)}{2 g}-\frac{E\left(\left.\sin ^{-1}(b)\right|g\right)}{2 (g-1) g}$

The other derivatives work also.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.