65 lines
1.5 KiB
Python
65 lines
1.5 KiB
Python
|
import numpy as np
|
||
|
|
import matplotlib.pyplot as plt
|
||
|
|
|
||
|
|
class StochastikIterator(object):
|
||
|
|
def __init__(self, r, num_simulations, chi0, std_deviation, maxiter=None):
|
||
|
|
self._r = r
|
||
|
|
self._num_simulations = num_simulations
|
||
|
|
self._chi0 = chi0
|
||
|
|
self._std_deviation = std_deviation
|
||
|
|
self._maxiter=maxiter
|
||
|
|
self._chi = None
|
||
|
|
self._currentiter = None
|
||
|
|
|
||
|
|
def __iter__(self):
|
||
|
|
self._currentiter = 0
|
||
|
|
self._chi = np.array([self._chi0] * self._num_simulations, dtype=np.float64)
|
||
|
|
return self
|
||
|
|
|
||
|
|
def __next__(self):
|
||
|
|
if(self._maxiter and self._currentiter == self._maxiter):
|
||
|
|
raise StopIteration()
|
||
|
|
self._currentiter += 1
|
||
|
|
|
||
|
|
|
||
|
|
result = self._chi.copy()
|
||
|
|
self._chi *= np.exp(r - np.random.normal(scale=self._std_deviation, size=self._num_simulations))
|
||
|
|
|
||
|
|
return result
|
||
|
|
|
||
|
|
|
||
|
|
|
||
|
|
def do(self, iters=None):
|
||
|
|
if(not (iters or self._maxiter)):
|
||
|
|
raise ValueError("iters or maxiter must be specified")
|
||
|
|
|
||
|
|
if(iters):
|
||
|
|
until = iters
|
||
|
|
if(self._maxiter and (not iters or (iters and self._maxiter < iters))):
|
||
|
|
until = self._maxiter
|
||
|
|
|
||
|
|
|
||
|
|
self._currentiter = 0
|
||
|
|
self._chi = np.array([self._chi0] * self._num_simulations, dtype=np.float64)
|
||
|
|
for i in range(until):
|
||
|
|
self._chi *= np.exp(r - np.random.normal(scale=self._std_deviation, size=self._num_simulations))
|
||
|
|
|
||
|
|
return self._chi.copy()
|
||
|
|
|
||
|
|
|
||
|
|
|
||
|
|
mean_chi = lambda r: np.mean(StochastikIterator(r, 500, 100, 0.25).do(70))
|
||
|
|
|
||
|
|
for r in [0.05, 0, -0.05]:
|
||
|
|
print("r =", r, "\t<chi> = ", mean_chi(r))
|
||
|
|
|
||
|
|
|
||
|
|
R = np.arange(-4, 4, 0.01)
|
||
|
|
|
||
|
|
M = [mean_chi(r) for r in R]
|
||
|
|
|
||
|
|
plt.plot(R, M, label="$<\\chi>$")
|
||
|
|
plt.show()
|
||
|
|
|
||
|
|
|