Handling p0 input in source

[djps]

Describe the bug The logic of defining an initial pressure distribution is different from all other methods in that it can not be modified after being defined, throwing the error: TypeError: 'NoneType' object does not support item assignment for something like

source.p0 = np.zeros((Nx,Ny,Nz))
source.p0[Nx // 2 -1, Ny // 2 -1, Nz // 2 -1] = 1

This would be fine for ux etc. There are ways to handle it, i.e.

p0 = np.zeros((Nx,Ny,Nz))
p0[Nx // 2 -1, Ny // 2 -1, Nz // 2 -1] = 1
source.p0 = p0

but the logic is confusing.

[djps]

It appears that https://github.com/waltsims/k-wave-python/blob/d675e2d192c96128f5755c8d17fcc5d77a691b7d/kwave/ksource.py#L62 has np.sum(val !=0) which should be replaced by a statement with np.all() statement equivalent, as the setter overwrites things to None when it shouldn't.