Overlapping atoms with open boundary cartesian coordinate unit cells

[drjbarker]

Robert is making large unit cells defined in cartesian coordinates. The basis vectors look like

    basis = (                         // unit cell
    [220.808, 0.0, 0.0],
    [0.0, 220.808, 0.0],
    [0.0, 0.0, 39.0335]);

Atom positions such as

O 108.73350799 220.80799978 1.951675 
Sr 108.73350799 0.0 1.951675

are then found to be too close together in JAMS which then fails with an error.

I think this is because the cartesian coordinates are transformed to fractional coordinates internally on [0,1) which then puts these two atoms very close together. For open boundary systems this is confusing and we could try to avoid the error somehow or at least give more useful information.