Feature/distributions part2

[whaeck]

This PR adds the following one dimensional function components:

• Legendre and tabulated angular distributions
• tabulated energy distributions

This PR adds the following two dimensional function components:

• Legendre angular distribution data on a grid
• Tabulated angular distribution data on a grid
• Tabulated energy distribution data on a grid

This is based on the latest scion developments

[cjosey]

[Questions]
Are there no equivalent Python tests for the energy-angle evaluation?
Do operators like chunk(energy, angle) work with numpy.ndarray inputs or do you have to construct an numpy.vectorize object?

Otherwise, looks good.

Edit: I see now, the Python interface is in part 3.

[whaeck]

chunk(energy, angle) is a function call that interpolates on the the data.

[cjosey]

chunk(energy, angle) is a function call that interpolates on the the data.

I meant more in this sense:

import numpy as np

e_vals = np.logspace(-5, 8, 100)
mu_vals = np.linspace(-1, 1, 50)

e_mesh, mu_mesh = np.meshgrid(e_vals, mu_vals)

val = chunk(e_mesh, mu_mesh)

If it's only a scalar argument, I could just do

chunk_vecfun = np.vectorize(chunk)

val = chunk_vecfun(e_mesh, mu_mesh)

[whaeck]

Oh, that's an interesting use case.

The function is specifically only for scalar arguments, and only returns a double. I won't be able to do it with the call operator - as C++ requires the same return type when using operator overloading. Still, I'll look into it.