Make a function closure to implement a chemical Langevin (continuous-state) approximation for a SPN.
Arguments
- S
a stoichiometry
Matrix-classobject- Sout
an optional matrix to track of event firings. In the continuous stochastic model this will be the approximate cumulative intensity of each event.
- haz
a list of hazard functions
- dt
time-step for Euler-Maruyama method used to solve the SDE system
- maxhaz
maximum allowable hazard
Value
function closure for use in sim_trajectory_R or sim_trajectory_CSV
Details
The chemical Langevin approximation is a numerical simulation of a Fokker-Planck approximation to the Master equations (Kolmogorov Forwards Equations) governing the stochastic model; the CLE approximation is a second-order approximation that will get the correct mean and variance but higher order moments will be incorrect.
The design of step_CLE is from: Wilkinson, D. J. (2011). Stochastic
modeling for systems biology. CRC press
Elements of the N list come from two places: The stoichiometry matrix
(S) is generated in spn_S and the hazards (h) come
from spn_hazards.