Simulation

parse_model(network::Network)

Construct a state <: Vector{Int} and model <: ReactionSystem from a Network object.

Use this in simulate to avoid constructing internal data structures over multiple samples.

simulate(network::Network, algname::SimulationAlgorithm;
            [tfinal = 0.0],
            [rates_cache = HasRates],
            [save_points = nothing],
            [save_function = __extract])

Simulate a Network of interacting populations. Note that simulations may terminate early if the cumulative intensity reaches 0 (that is, reaching an absorbing state).

Keyword Arguments

• tfinal: The final simulation time.
• rates_cache: Indicates the type of information stored in a rates cache. If HasRates is chosen, store the actual rates. If HasSums is chosen, store partial sums of the rates. This effectively toggles between linear and binary searches, provided the algorithm supports the option.
• save_points: An indexable collection indicating time points to sample and record system state. The default nothing forces saving after every reaction event.
• save_function: A three argument function that maps the three arguments (simulator, state, model) to data recorded in a SamplePath. Default behavior is to record each species.

simulate(state, model, algname::SimulationAlgorithm;
            [tfinal = 0.0],
            [rates_cache = HasRates],
            [save_points = nothing],
            [save_function = __extract])

Simulate a model with the given initial state. Note that simulations may terminate early if the cumulative intensity reaches 0 (that is, reaching an absorbing state).

Arguments

For well-mixed systems:

• state: A Vector{Int} constructed by parse_model.
• model: A ReactionSystem type constructed by parse_model.

For lattice-based systems:

• state: A Lattice type whose topology comptabile is compatible with model.
• model: An InteractingParticleSystem type built from @enumerate_with_sclass.

Keyword Arguments

• tfinal: The final simulation time.
• rates_cache: Indicates the type of information stored in a rates cache. If HasRates is chosen, store the actual rates. If HasSums is chosen, store partial sums of the rates. This effectively toggles between linear and binary searches, provided the algorithm supports the option.
• save_points: An indexable collection indicating time points to sample and record system state. The default nothing forces saving after every reaction event.
• save_function: A three argument function that maps the three arguments (simulator, state, model) to data recorded in a SamplePath. Default behavior is to store population counts for well-mixed systems or Configuration objects for lattice-based systems.