I believe a number of factors in the edge integration of transfer functions are not actually necessary in how we do the integration. This is part motivated by maths (don't need the Jacobian), but largely by experimentation -- the transfer functions seem to behave better for large h when a number of factors are dropped.
Also a few typo corrections in the about sections.
I believe a number of factors in the edge integration of transfer functions are not actually necessary in how we do the integration. This is part motivated by maths (don't need the Jacobian), but largely by experimentation -- the transfer functions seem to behave better for large
hwhen a number of factors are dropped.Also a few typo corrections in the about sections.