Spectrally accurate 1D indefinite integration

[mrhardman]

A note to record that the present calculation of phi from Epar for the kinetic and fluid electron options

moment_kinetics/moment_kinetics/src/em_fields.jl

Line 177 in eac094f

function calculate_phi_from_Epar!(phi, Epar, r, z)

uses a (presumably secord-order accurate) trapezoidal rule. It is possible to formulate a 1D indefinite integral using the same spectral methods as used to form the differentiation matrices in moment_kinetics, with the caveat that there is a free integration constant to be accounted for. If the accuracy of the 1D indefinite integration method is a limiting factor in the calculation for any user/developer, please comment here and I will implement the spectrally accurate indefinite integration method.

[mrhardman]

I have made an initial start on a spectrally accurate 1D line integration feature here https://github.com/mabarnes/moment_kinetics/tree/line-integration.

[mrhardman]

Now addressed in #299