I am thinking how to deduct the mean gradient in polar coordinates if the process itself is Poisson/Gaussian/... The basic deduction of the gradient is in the thread How to obtain the gradient in polar coordinates so the gradient operator is
\begin{equation} \nabla =\hat{\pmb e}_{r}\frac{\partial }{\partial r}+\hat{\pmb e}_{\theta}\frac{1}{r}\frac{\partial }{\partial \theta}. \end{equation}
I am thinking how I should proceed here. I am monitoring confounders of multifactorial processes. I can plot many of them separately but I would like characterise an individual process by a mean vector.
