Hi,
I might be mistaken here, bit I think the theory page (https://github.com/maroba/findiff/blob/master/docs/source/theory.rst) contains a typo in the Kronecker delta.
To give the desired result of $fk^{(n)}$ it should be $\delta{n \alpha}$ instead of $\delta_{k \alpha}$ (which would give $f_k^{(k)}$).
(Also, do you have a reference regarding the claim made in the last line regarding optimal stencils in higher dimensions being the sum of 1D stencils?)
Hi, I might be mistaken here, bit I think the theory page (https://github.com/maroba/findiff/blob/master/docs/source/theory.rst) contains a typo in the Kronecker delta. To give the desired result of $fk^{(n)}$ it should be $\delta{n \alpha}$ instead of $\delta_{k \alpha}$ (which would give $f_k^{(k)}$).
(Also, do you have a reference regarding the claim made in the last line regarding optimal stencils in higher dimensions being the sum of 1D stencils?)
Best regards