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 = 3 sizeStates = 4 sizeConstants = 9 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] = "x in component x (per_mm3)" legend_constants[0] = "lamda in component x (per_mm3_per_day)" legend_constants[1] = "d in component x (per_day)" legend_constants[2] = "beta in component kinetic_parameters (mm3_per_day)" legend_algebraic[0] = "s in component kinetic_parameters (dimensionless)" legend_states[1] = "y in component y (per_mm3)" legend_constants[3] = "a in component y (per_day)" legend_constants[4] = "p in component kinetic_parameters (mm3_per_day)" legend_states[2] = "z in component z (per_mm3)" legend_algebraic[1] = "log_y in component y (dimensionless)" legend_states[3] = "w in component w (per_mm3)" legend_constants[5] = "b in component w (per_day)" legend_constants[6] = "c in component kinetic_parameters (mm3_mm3_per_day)" legend_constants[7] = "q in component kinetic_parameters (per_mm3)" legend_algebraic[2] = "log_w in component w (dimensionless)" legend_constants[8] = "h in component z (per_day)" legend_rates[0] = "d/dt x in component x (per_mm3)" legend_rates[1] = "d/dt y in component y (per_mm3)" legend_rates[3] = "d/dt w in component w (per_mm3)" legend_rates[2] = "d/dt z in component z (per_mm3)" return (legend_states, legend_algebraic, legend_voi, legend_constants) def initConsts(): constants = [0.0] * sizeConstants; states = [0.0] * sizeStates; states[0] = 10 constants[0] = 1 constants[1] = 0.1 constants[2] = 0.5 states[1] = 0.1 constants[3] = 0.2 constants[4] = 1 states[2] = 0 states[3] = 0.001 constants[5] = 0.01 constants[6] = 0.1 constants[7] = 0.5 constants[8] = 0.1 return (states, constants) def computeRates(voi, states, constants): rates = [0.0] * sizeStates; algebraic = [0.0] * sizeAlgebraic rates[3] = constants[6]*states[0]*states[1]*states[3]-(constants[6]*constants[7]*states[1]*states[3]+constants[5]*states[3]) rates[2] = constants[6]*constants[7]*states[1]*states[3]-constants[8]*states[2] algebraic[0] = custom_piecewise([less_equal(voi , 15.0000), 1.00000 , greater_equal(voi , 40.0000), 1.00000 , True, 0.00420000]) rates[0] = constants[0]-(constants[1]*states[0]+algebraic[0]*constants[2]*states[0]*states[1]) rates[1] = algebraic[0]*constants[2]*states[0]*states[1]-(constants[3]*states[1]+constants[4]*states[1]*states[2]) return(rates) def computeAlgebraic(constants, states, voi): algebraic = array([[0.0] * len(voi)] * sizeAlgebraic) states = array(states) voi = array(voi) algebraic[0] = custom_piecewise([less_equal(voi , 15.0000), 1.00000 , greater_equal(voi , 40.0000), 1.00000 , True, 0.00420000]) algebraic[1] = log(states[1]*1.00000, 10) algebraic[2] = log(states[3]*1.00000, 10) return algebraic def custom_piecewise(cases): """Compute result of a piecewise function""" return select(cases[0::2],cases[1::2]) 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)