Yes, it's possible that a simulation would use a non-uniform lattice-like data structure. Maybe something like a quad-tree (with appropriate dimensionality) whe

Yes, it's possible that a simulation would use a non-uniform lattice-like data structure. Maybe something like a quad-tree (with appropriate dimensionality) where the granularity of the structure varies according to the local density of data. Granted, such a data structure would need to be expressed in a coordinate system, which itself defines a grid or matrix. But can't coordinate systems use non-uniform representations? (analogous to floating-point) Am neither physicist nor CS person, so not sure all of this holds together, just wondering.