Can there exist a $C^2$ function $f: \mathbb R^3 \to \mathbb R$ with $f_x = 3x - 2y + 4z, f_y = -2x+3y +z $ and $f_z = 3x + y - 5z$. Explain your answer.

$\int f_udu = \frac{3}{2}x^2 - 2xy + 4zu + f(y, z) = f(x, y, z)$

$\int f_y dy = -2xy + zy + \frac{3}{2}y^2 + f(x, z) = f(x, y, z)$

$\int f_z dz = 3xz + yz - \frac{5z^2}{2} + f(x, y) = f(x, y, z)$

since $\frac{d}{dz} (\int f_u du) \neq f_z$ hence does not exist any for which the partial derivatives are given

is this correct?


There's a few different approaches. Yours is fine.

Hans Engler has a really good one. Here's a third, somewhere in between.

You can make a vector $\vec{v}=<f_x,f_y,f_z>$

Then if $\nabla \times \vec{v}=0$ then such a function exists.

$\nabla \times \vec{v}\implies v_i=\sum_{j=1}^3\sum_{k=1}^3 \epsilon_{ijk}\frac{\partial}{\partial x^j}\frac{\partial f}{\partial x_k}$

Where $\epsilon_{ijk}$ is the Levi-Civita symbol. $x_1=x, x_2=y, x_3=z.$

You recover Mr. Engler's with perhaps a bit more of work.

Through abuse of notation you can find $\nabla \times \vec{v}$, the curl of $\vec{v}$ with a matrix.


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.