def __init__(self):
fitglobals.debugoff()
P = noise.NoiseParams()
P.tstop = 20
P.dt = 0.1
A0 = 0.5
A1 = 0.1
P.A = numpy.matrix([[A0, 0], [0, A1]])
# P.B, next line define noise injected to each component, uncorrelated
P.B = numpy.matrix([[0.5, 0], [0, 0.4]])
P.InitialCov = numpy.matrix([[1, 0], [0, 1]])
elist = numpy.arange(0.1, 20, 0.1).tolist()
elist2 = numpy.arange(0.05, 20.0, 0.05).tolist()
O1 = models.ObserveState0(P, 5)
O2 = models.ObserveStateSum(P, 6)
# O1.Times.set([2,4,6,8,10,12,14,16,18,20])
# O2.Times.set([3,6,9,12,15,18])
O1.Times.set(elist)
O2.Times.set(elist)
O1.sigma = 0.001
O2.sigma = 0.0001
Obs = models.ObservationModel(P, 5, [O1, O2])
Sys = models.DecayModel(P, 0, 2)
# Sys.Injection.set([1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20])
Sys.Injection.set(elist2)
Initial = numpy.matrix([[1.0], [2.0]])
self.M = models.Model(Sys, Obs, P, Initial)
self.sim(False)
def __init__(self):
P = noise.NoiseParams()
P.tstop = 20
P.dt = 0.1
P.A = numpy.matrix(0.5)
P.B = numpy.matrix([0.1])
P.InitialCov = numpy.matrix(1)
O1 = models.ObserveState0(P, 5)
elist = numpy.arange(0.1, 20, 0.1).tolist()
elist2 = numpy.arange(0.05, 20.0, 0.05)
# O1.Times.set([1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20])
O1.Times.set([2, 4, 6, 8, 10, 12, 14, 16, 18, 20])
# O1.Times.set(elist)
O1.sigma = 0.001
Obs = models.ObservationModel(P, 5, [O1])
Sys = models.DecayModel(P, 0, 1)
# Sys.Injection.set(elist2)
Sys.Injection.set([
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,
20
])
self.M = models.Model(Sys, Obs, P, numpy.matrix(2))
self.sim(False)
fitglobals.debugon() P = noise.NoiseParams() P.tstop = 20 P.dt = 0.1 P.A = numpy.matrix(0.5) P.B = numpy.matrix([0.5, 0.4, 0.3]) P.InitialCov = numpy.matrix(1) O1 = models.ObserveState0(P,5) O2 = models.ObserveState0(P,6) O1.Times.set([2,4,6,8,10,12,14,16,18,20]) O2.Times.set([3,6,9,12,15,18]) O1.sigma = 0.001 O2.sigma = 0.0001 Obs = models.ObservationModel(P,5,[O1,O2]) #Sys = models.DecayModel(P,0,3) Sys = models.NeuronModel(P,0,3) Sys.Injection.set([1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20]) M = models.Model(Sys,Obs,P) # Turn noise off and simulate P.B = numpy.matrix([0.,0.,0.]) # Not needed: change(P) M.Obs.C[0].sigma = 0 M.Obs.C[1].sigma = 0 Data = M.sim() # Turn noise back on and calculate log-likelihood P.B = numpy.matrix([0.5, 0.4, 0.3]) # Not needed: change(P)