Nexus: Adding support for ghost atoms

[shivupa]

Allows nexus to understand ghost atoms.

Why this is necessary:
The converters from quantum chemistry codes (e.g. gaussian/gamess) don't currently know about ghost atoms either, but they will parse the correct basis functions/MO coefficients. These just get assigned to atoms of type "". Replacing these with something reasonable like "X" and trying to uses nexus post calculation for analysis results in

PhysicalSystem error:
    particle X is unknown
  exiting.

This fixes the above error so now QmcpackAnalyzer can be used on systems with ghost atoms.

[qmc-robot]

Can one of the maintainers verify this patch?

[markdewing]

okay to test

[jtkrogel]

This looks like it should work, though permanently adding atom X to the periodic table is unusual.

Does the following work?

from physical_system import Matter,Ion
from qmcpack_analyzer import QmcpackAnalyzer

Matter.particle_collection.add_particles(Ion('X'))

qa = QmcpackAnalyzer('./your.in.xml')
qa.analyze()

[shivupa]

Matter.particle_collection.add_particles(Ion('X',mass=0,charge=0,spin=0,protons=0,neutrons=0))

This does work. I could use this since I'm guessing ghost atoms aren't very commonly used.

[jtkrogel]

I think having a way of supporting ghost atoms is a good idea.

One possibility is to add a function like this to physical_system.py

def ghost_atoms(*names):
    for name in names:
        Matter.particle_collection.add_particles(Ion(name,mass=0,charge=0,spin=0,protons=0,neutrons=0))
    #end for
#end def ghost_atoms

and use it outside

ghost_atoms('X')
qa = QmcpackAnalyzer(...)

or move its use into QmcpackAnalyzer instead

qa = QmcpackAnalyzer(...,ghost_atoms=['X'])

If having an atom X permanently available is more convenient, then I'm fine if you want to add this line

Matter.particle_collection.add_particles(Ion('X',mass=0,charge=0,spin=0,protons=0,neutrons=0))

after Matter.set_particle_collection(Particles(plist)) in physical_system.py. This way 'X' is always there as a vacuous ion, but it is not in the periodic_table which is queried for real data.

[shivupa]

That makes more sense, and this way ghost atoms don't have to be named "X". I'll update this PR.

[shivupa]

@jtkrogel How does this look? I didn't love having to pop from kwargs since it is mutating the object, but it was necessary otherwise the __init__ for QmcpackAnalysisRequest would break.

[jtkrogel]

Just one small requested change, then good to go.

[shivupa]

Fixed!

[jtkrogel]

Looks good. Thanks.