Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I have a question about, how can i derivate an expression of the form $$ F\left( u \right) = \int\nolimits_{C_1 }^{C_2 } {f\left( {ux} \right)\mathrm{d}x} $$ where $ C_1 ,C_2 $ are constants. I have no idea :/

share|cite|improve this question
This is called differentiating under the integral sign. But since $C_1$ and $C_2$ are constants, the answer is simply $\int_{C_1}^{C_2} \frac{\partial}{\partial u}f(ux) dx$. – Srivatsan Sep 4 '11 at 14:14
..., which is $\int_{C_1}^{C_2}xf'(ux)\mathrm dx=\frac1u\left([xf(ux)]_{C_1}^{C_2}-\int_{C_1}^{C_2}f(ux)\mathrm dx\right)$ (where the last step is only valid for $u\neq0$). – joriki Sep 4 '11 at 14:30
But for example here $$ \eqalign{ & f\left( a \right) = \int\limits_0^\infty {\frac{{Arc\tan \left( {ax} \right) - Arc\tan \left( x \right)}} {x}dx} \cr & \cr & \frac{d} {{da}}f\left( a \right) = \int\limits_0^\infty {\frac{{dx}} {{1 + a^2 x^2 }}} \cr} $$ He forgets the 1/x in the expression he only considered Arctan(ax) but I think that he should derivate with respect to a $ Arctan (ax) / x $ )= – August Sep 4 '11 at 14:32
@August, The new example is not of the form you gave in your question. And the differentiation seems ok. Note that you are (partial) differentiating w.r.t. $a$, not $x$. So the $x$ is a constant. Also differentiating $\arctan (ax)$ w.r.t. $a$ gives a spare $x$, thanks to chain rule. These two $x$'s cancel nicely. – Srivatsan Sep 4 '11 at 14:56
up vote 4 down vote accepted

This operation is called differentiating under the integral sign. In the general case, $$ F(u) = \int_{a(u)}^{b(u)} g(u,x) dx, $$ where the limits are functions of the parameter $u$, we have $$ F'(u) = f(u, b(u)) b'(u) - f(u, a(u)) a'(u) + \int_{a(u)}^{b(u)} \frac{\partial}{\partial u} g(u,x) dx. $$ In your case, the limits are constants, and $g(u,x)$ has the special form $f(ux)$. So the first two terms drop out, giving: $$ F'(u) = \int_{C_1}^{C_2} \frac{\partial}{\partial u} f(ux) dx = \int_{C_1}^{C_2} f'(ux) x dx, $$ thanks to the chain rule.

In general, one cannot "simplify" such answers any further, but in your case you can. Noting that $$ F'(u) = \int_{C_1}^{C_2} \frac{x}{u} \frac{\partial}{\partial x} f(ux) dx, $$ and integrating by parts, we get: $$ F'(u) = \left. \frac{x}{u} f(ux) \right|_{C_1}^{C_2} - \int_{C_1}^{C_2} \left(\frac{\partial}{\partial x} \frac{x}{u} \right) \cdot f(ux) dx = \ldots $$ I will leave it to you to complete the answer. You should be able to express the final integral in terms of $F(u)$ itself.

Edit: As @joriki notes in a comment, the "simplification" step is valid only if $u \neq 0$.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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