initial version of a VelocityBoltzmann with centre of mass shift

[reinimaurer1]

Adds an optional keyword argument so that VelocityBoltzmann can be set with a center of mass velocity. This is useful for atomic projectiles and low-dimensional models. (Diatomics are already dealt with within the DiatomicQuantised initial conditions)

@clbox is this functionality that you might have already implemented in your yet-to-be-merged edits for atomic scattering?

still playing around with the behavior. Currently, I have only tested the 1D case. Looking for a quick 3D case to make sure it works. Any ideas?

[Alexsp32]

I've shortened the centre definition as James suggested and added the capability to modify Distribution centres with + and - after generating them as well.

e.g.

dist_3d=VelocityBoltzmann(300, fill(2000, (3,4)), (3,4))
centre_of_mass_translation=[1,0,0] # centre of mass velocity vector
shift_matrix=repeat(centre_of_mass_translation; outer=[1,4]) #repeat translation vector for each atom
shifted_dist=dist_3d+shift_matrix #returns as if centre were provided on generating Boltzmann distribution

[reinimaurer1]

I like the behaviour that you implemented @Alexsp32 ! Looking at this, I don't think that my optional keyword is even needed anymore and I can revert my edits to VelocityBoltzmann distribution. Would you agree?

[jamesgardner1421]

It's probably better to define a new function for this operation instead of overloading +. This is what they call "type piracy" where you implement methods for functions that you don't own on types that you don't own.

[Alexsp32]

Good point. In that case, I'd just remove that function without offering an alternative, since it's easier to re-initialise a single normal distribution with a new $\mu$ than to do it for an entire Matrix.

[reinimaurer1]

this version should do for now. It certainly does what it is meant to do.