Cross derivatives of a multivariate function?

Hello,

I hope all is well. I have a question about the title. I am interested in computing a $M\times N \times N$ three-way array of derivatives:

$\frac{\partial^{2}f_{i}}{\partial x_{j}\partial x_{k}}$

where $f$ is a $M$ vector of functions $f_i \ (i=1,\ldots,M)$, and each $f_i$ takes $x \in \mathbb{R}^N$ as its arguments. What is the best way to compute this object in FiniteDifferences.jl?

This is clearly not a Jacobian, but I tried to do jacobian(central_fdm(5,2), myfun, myval) just to what it gives. It did not give an error or an $M\times N \times N$ array but an $M\times N$ array. So what is it supposed to compute?

Best,
Daisuke

Start with the public jacobian and central_fdm entry points used in the issue, and check how they handle the vector-valued myfun. Determine whether the requested M×N×N cross-derivative object is supported; done means the supported usage or limitation is clearly documented.