1
$\begingroup$

I'm reading a textbook where it forms a Lagrangian function

$$ L = \int_0^1 f(x)^{1 - \frac{1}{\alpha}}dx - \lambda\int_0^1 g(x) f(x) dx$$ But how do you take the derivative of this thing? The variable that we want to optimize is $f(x)$, which is a function ($g(x)$ is a known function). So we need to take the derivative with respect to a function? How?

So in this case, what are the first-order conditions?

$\endgroup$
3
$\begingroup$

An important result of the calculus of variations is that if you have a functional $L[f(x)]$ such that $$ L[f] =\int_a^b J(x,f,f') dx, $$ $L$ is minimized if $$ \frac{\partial J}{\partial f} - \frac{d}{dx} \frac{\partial J}{\partial L'} = 0. $$ This is the Euler-Lagrange equation. In your case, $$ L[f] = \int_0^1 \left[f(x)^{1-1/\alpha}-\lambda g(x) f(x) \right] dx, $$ therefore the Euler-Lagrange equation leads to $$ \frac{\partial}{\partial f} \left[f(x)^{1-1/\alpha}-\lambda g(x) f(x) \right] - \frac{d}{dx} \frac{\partial}{\partial f'} \left[f(x)^{1-1/\alpha}-\lambda g(x) f(x) \right] =0. $$ The derivative in relation to $f'$ vanishes because $L[f]$ does not depend on $f'$. Evaluating the derivative in relation to $f$: $$ \frac{\alpha-1}{\alpha}f(x) ^{-1/\alpha} - \lambda g(x) = 0 $$ and now solving for $f$: $$ f(x) = \left( \frac{\lambda \alpha}{\alpha-1} g(x)\right)^{-\alpha}, $$ which is the $f(x)$ that minimize your $L$.

$\endgroup$

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.