Generated Code
The following is python code generated by the CellML API from this CellML file. (Back to language selection)
The raw code is available.
# Size of variable arrays: sizeAlgebraic = 1 sizeStates = 4 sizeConstants = 12 from math import * from numpy import * def createLegends(): legend_states = [""] * sizeStates legend_rates = [""] * sizeStates legend_algebraic = [""] * sizeAlgebraic legend_voi = "" legend_constants = [""] * sizeConstants legend_voi = "time in component environment (day)" legend_states[0] = "T in component T (dimensionless)" legend_constants[0] = "k in component T (first_order_rate_constant)" legend_constants[1] = "r in component T (first_order_rate_constant)" legend_constants[2] = "d in component T (first_order_rate_constant)" legend_constants[3] = "gamma in component T (first_order_rate_constant)" legend_states[1] = "C in component C (dimensionless)" legend_states[2] = "A in component A (dimensionless)" legend_constants[4] = "lambda in component A (first_order_rate_constant)" legend_constants[5] = "delta_1 in component A (first_order_rate_constant)" legend_constants[6] = "alpha in component kinetic_parameters (first_order_rate_constant)" legend_states[3] = "A_star in component A_star (dimensionless)" legend_constants[7] = "delta_2 in component A_star (first_order_rate_constant)" legend_constants[8] = "eta in component C (first_order_rate_constant)" legend_constants[9] = "epsilon in component C (dimensionless)" legend_constants[10] = "q in component C (first_order_rate_constant)" legend_constants[11] = "mu in component C (first_order_rate_constant)" legend_algebraic[0] = "R in component ratio (dimensionless)" legend_rates[0] = "d/dt T in component T (dimensionless)" legend_rates[2] = "d/dt A in component A (dimensionless)" legend_rates[3] = "d/dt A_star in component A_star (dimensionless)" legend_rates[1] = "d/dt C in component C (dimensionless)" return (legend_states, legend_algebraic, legend_voi, legend_constants) def initConsts(): constants = [0.0] * sizeConstants; states = [0.0] * sizeStates; states[0] = 0.1 constants[0] = 10 constants[1] = 0.5 constants[2] = 0.1 constants[3] = 1 states[1] = 0.015 states[2] = 1 constants[4] = 1 constants[5] = 0.1 constants[6] = 0.05 states[3] = 2 constants[7] = 1.5 constants[8] = 2 constants[9] = 1 constants[10] = 0.5 constants[11] = 0.1 return (states, constants) def computeRates(voi, states, constants): rates = [0.0] * sizeStates; algebraic = [0.0] * sizeAlgebraic rates[0] = (constants[1]*states[0]*(1.00000-(states[0]*1.00000)/constants[0])-constants[2]*states[0])-constants[3]*states[0]*states[1] rates[2] = (constants[4]-constants[5]*states[2])-constants[6]*states[2]*states[0] rates[3] = constants[6]*states[2]*states[0]-constants[7]*states[3] rates[1] = ((constants[8]*states[3]*states[1])/(constants[9]*states[1]+1.00000)-constants[10]*states[0]*states[1])-constants[11]*states[1] return(rates) def computeAlgebraic(constants, states, voi): algebraic = array([[0.0] * len(voi)] * sizeAlgebraic) states = array(states) voi = array(voi) algebraic[0] = (states[1]*states[3])/(constants[10]*1.00000*states[0]) return algebraic def solve_model(): """Solve model with ODE solver""" from scipy.integrate import ode # Initialise constants and state variables (init_states, constants) = initConsts() # Set timespan to solve over voi = linspace(0, 10, 500) # Construct ODE object to solve r = ode(computeRates) r.set_integrator('vode', method='bdf', atol=1e-06, rtol=1e-06, max_step=1) r.set_initial_value(init_states, voi[0]) r.set_f_params(constants) # Solve model states = array([[0.0] * len(voi)] * sizeStates) states[:,0] = init_states for (i,t) in enumerate(voi[1:]): if r.successful(): r.integrate(t) states[:,i+1] = r.y else: break # Compute algebraic variables algebraic = computeAlgebraic(constants, states, voi) return (voi, states, algebraic) def plot_model(voi, states, algebraic): """Plot variables against variable of integration""" import pylab (legend_states, legend_algebraic, legend_voi, legend_constants) = createLegends() pylab.figure(1) pylab.plot(voi,vstack((states,algebraic)).T) pylab.xlabel(legend_voi) pylab.legend(legend_states + legend_algebraic, loc='best') pylab.show() if __name__ == "__main__": (voi, states, algebraic) = solve_model() plot_model(voi, states, algebraic)