使用司仪将模型拟合到实验数据

Question

我正在尝试将以下ODE模型与实验数据拟合： 当前，我正在使用emcee查找适合模型的参数。以下代码是使用scipy.odeint的ODE模型。

class IFFL:
def __init__(self, m_production, m_degradation, s_production,
             s_degradation, ms_degradation, k_p, k_i, LacI):
    self.m_production = m_production #mf-Lon
    self.m_degradation = m_degradation
    self.s_production = s_production #m-Scarlet
    self.s_degradation = s_degradation
    self.ms_degradation = ms_degradation #m-Scarlet degradation by mf-Lon
    self.k_p = k_p #dissociation constant for LacI and pLac
    self.k_i = k_i #dissociation constant for IPTG and LacI
    self.LacI = LacI
    self.IPTG = []

def inducer(self, T, P, D):
    sw = np.zeros(len(T))
    iterate = 0
    sw[iterate] = 0
    iterate += 1
    while iterate < (len(T)):
        for i in range(int(D)):
            if iterate >= len(T):
                return sw
            sw[iterate] = 100000
            iterate += 1
        for i in range(int(P - D)):
            if iterate >= len(T):
                return sw
            sw[iterate] = 0
            iterate += 1
    return sw

def model(self, z, t, IPTG):
    x = z[0]
    r = z[1]
    dxdt = self.m_production/(1 + self.LacI/self.k_p/(1+(IPTG/self.k_i)**2)) - self.m_degradation*x #x dot
    drdt = self.s_production/(1 + self.LacI/self.k_p/(1+(IPTG/self.k_i)**2)) - self.ms_degradation*(x**2/(x**2 + 1))*r - self.s_degradation*r #r dot
    dzdt = [dxdt,drdt]
    return dzdt

def runmodel(self, m, s, t, P, D):
    z0 = [m,s]
    self.IPTG = self.inducer(t, P, D)
    print(self.IPTG)
    mr = np.empty_like(t)
    sr = np.empty_like(t)
    mr[0] = z0[0]
    sr[0] = z0[1]
    for i in range(1, len(t)):
        tspan = [t[i-1],t[i]]
        z = odeint(self.model,z0,tspan,args = (self.IPTG[i],), full_output = 0)
        mr[i] = z[1][0]/(self.m_production/self.m_degradation) #normalized
        sr[i] = z[1][1]/(self.s_production/self.s_degradation)
        z0 = z[1]
    #plt.plot(mr, sr) #uncomment for solution trajectory
    return [mr, sr]

长话短说，我有五个参数需要与实验数据拟合，这些参数是：

1. m_production和m_degradation
2. s_production和s_degradation
3. ms_degradation

我尝试遵循emcee的代码，但是看来我需要“最佳猜测”参数值。我该如何解决这个问题？