For a general function $g$, define $f=\mathbb{E}_{(x,y)\sim \mathcal{N}(0, \Sigma)}[g(x)g(y)]$, where $\Sigma=\begin{pmatrix}a & c \\c & b \end{pmatrix}$, and $a>0,b>0,ab>c^2$.

I am looking for a Mathematica solution for finding the partial derivative with respect to $c$ and evaluate it at the point $(1,1,0)$.

The solution by hand is $$ \frac{\partial f}{\partial c}|_{a=1,b=1,c=0}(\mathbb{E}_{z\sim\mathcal{N}(0,1)}[zg(z)])^2. $$

I've tried the code:

gg = Function[{a, b, c}, 
   g[m] g[n], {m, n} \[Distributed] 
    MultinormalDistribution[{0, 0}, {{a, c}, {c, b}}]]]

Derivative[0, 0, 1][gg][1, 1, 0]

but I got something that really confuses me. Thanks for your help in advance.


Your Answer

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

Browse other questions tagged or ask your own question.