Types of LGCA

The simulator implements several lattice-gas cellular automata families. The two main modelling axes are volume exclusion and identity tracking. A third axis, n_species, selects classical multi-species classes.

The snippets below assume from lgca import get_lgca.

Classical LGCA

Classical LGCA enforce volume exclusion: each channel is either empty or occupied by one particle. Particles do not carry individual properties. The state array stores boolean-like occupancy values with the channel axis last.

Use this family with:

lgca = get_lgca(ve=True, ib=False)

Identity-based LGCA

Identity-based LGCA also enforce volume exclusion, but occupied channels store particle IDs instead of plain occupancy. Per-particle properties are stored in lgca.props and can be recorded during simulations.

Use this family with:

lgca = get_lgca(ve=True, ib=True)

LGCA without volume exclusion

NoVE LGCA allow channels to hold counts greater than one. This is useful for density-based models where individual cells do not need to be distinguished. Several NoVE interactions assume no rest channels; pass restchannels=0 when using alignment-style NoVE rules such as dd_alignment or di_alignment.

Use this family with:

lgca = get_lgca(ve=False, ib=False, restchannels=0)

Identity-based LGCA without volume exclusion

NoVE identity-based LGCA allow multiple labelled cells per node while preserving individual or family properties. These models are used for heterogeneous proliferating populations, including the go-or-grow and glioblastoma interaction variants.

Use this family with:

lgca = get_lgca(ve=False, ib=True, interaction="go_or_grow")

Multi-species LGCA

Multi-species LGCA are selected with n_species > 1. They are currently classical models, with or without volume exclusion. Identity-based multi-species models are not implemented. Multi-species state arrays include a species axis before the channel axis.

Use this family with:

lgca = get_lgca(n_species=2, ve=True, restchannels=1,
                interaction="excitable_medium_ms")

The current multi-species interaction library is intentionally small. See Built-in interactions for supported rules.

Data model summary

Family	Channel content	Individual properties	Species axis
Classical	Occupancy	No	No
Identity-based	Particle ID	Yes	No
NoVE classical	Particle count	No	No
NoVE identity-based	Lists or arrays of particle IDs	Yes	No
Multi-species	Occupancy or count by species	No	Yes