Nick Papior
Nick Papior
> Nodes could work nicely for this 🙄 Ah yeah, good idea! :rocket: What is the overhead here? And can you tell it to not do it *all-over*? Only on...
I seem to recall your nodes would traverse ast, and take everything, what we only want is the inputs, and the output, all in between shouldn't be nodified.
nice, I'll play with this soon! Sounds like a perfect fit here!
> Thanks! I will try to look at it during this week. > > I think that all `norm2` from an orbital should sum to 2 (if the calculation is...
`norm2` is equivalent to $\langle \psi | \mathbf S | \psi \rangle$, so for $\psi$ being eigenstates it should always be 1 (otherwise there is an error! ;)). Even for...
> Ok, so the problem here is that `norm2` is sometimes negative, and therefore can't be used as the size directly. Does a negative `norm2` make sense? In that case...
> I have also seen that `eigenstate.norm2()` of an eigenstate coming from spin-orbit returns an array of complex numbers. Is the imaginary part meaningful or just an artifact? The imaginary...
Ah, yes, I see. I think these relations are useful: ```python norm = norm2(projection="orbital") assert np.allclose(norm.sum(-1), 1) # I *think* the below will also hold: assert np.all(np.absolute(norm)
> ... > The imaginary part is an artifact because it can be used on `State` objects (which doesn't necessarily needs to be an eigenstate. Let me correct myself... The...
@pfebrer is there something that needs to be done here before we make a release? I think `abs` would make sense in plots.