from tvb.simulator.lab import *
import numpy as np
import math
import matplotlib.pyplot as plt
from tvb.rateML.XML2model import RateML
[docs]
class regularRun:
def __init__(self, sim_length, g, s, dt, period, omega = 60, filename='connectivity_zerlaut_68.zip'):
# def __init__(self, sim_length, g, s, dt, period, omega = 60, filename='paupau.zip'):
self.sim_length = sim_length
self.g = np.array([g])
self.s = np.array([s])
self.dt = dt
self.period = period
self.omega = omega * 2.0 * math.pi / 1e3
(self.connectivity, self.coupling) = self.tvb_connectivity(filename)
self.SC = self.connectivity.weights
[docs]
def tvb_connectivity(self, filename):
white_matter = connectivity.Connectivity.from_file(source_file=filename)
white_matter.configure()
white_matter.speed = np.array([self.s])
# white_matter_coupling = coupling.Linear(a=self.g)
# white_matter_coupling = coupling.Scaling(a=self.g)
white_matter_coupling = coupling.Coupling()
return white_matter, white_matter_coupling
[docs]
def tvb_python_model(self, modelExec):
# populations = models.Generic2dOscillator() # original
# populations = models.KuramotoT() # generated
# populations = models.OscillatorT() # generated
# populations = models.MontbrioT() # generated
# populations = models.RwongwangT() # generated
# populations = models.EpileptorT() # generated
model = 'models.' + modelExec + '()'
populations = eval(model)
populations.configure()
populations.omega = np.array([self.omega])
return populations
[docs]
def simulate_python(self, modelExec):
# Initialize Model
model = self.tvb_python_model(modelExec)
# zerlaut setup
noises = noise.Additive(nsig=np.array([0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0]), ntau=0.0)
integrator = integrators.HeunStochastic(dt=.1, noise=noises)
# Initialize integrator
# Initialize Monitors
monitorsen = (monitors.TemporalAverage(period=self.period))
# Initialize Simulator
sim = simulator.Simulator(model=model, connectivity=self.connectivity,
coupling=coupling.Linear(a=np.array(self.g), b=np.array(0.0)),
integrator=integrator,
monitors=[monitorsen])
sim.configure()
# sim.history.buffer[:] = 0.0
# sim.current_state[:] = 0.0
# print('shb', sim.history.buffer.shape) # ('n_time', 'n_cvar', 'n_node', 'n_mode')
(time, data) = sim.run(simulation_length=self.sim_length)[0]
# pad some zeros to make it equivalent to GPU for comparison
# CPU is now +1 timestep longer
# data = np.insert(data, 0, 0, axis=0)
# data = data[:-1]
# print('ds',data.shape)
plt.plot((data[:, 0, :, 0]), 'k', alpha=.2)
plt.show()
return (time, data)
if __name__ == '__main__':
# model_filename = 'Oscillator'
# language='python'
#
# RateML(model_filename, language)
simtime = 2000
g = .4
# g = 0.0042
s = 4.0
dt = .1
period = 10
model = 'Zerlaut_adaptation_second_order'
(time, data) = regularRun(simtime, g, s, dt, period).simulate_python(model)