def S(self):
if not hasattr(self, "_S"):
_S = []
for i, p in enumerate(self.P):
_S.append(steady_state(p))
# if np.array(_S).dtype is np.dtype('O'):
self._S = np.asarray(_S)
return self._S
def get_steady_state(tp):
"""Given a transition probability matrix, use ergodic.steady_state
to calculate the long-run steady-state, which is a vector
representing how long the system will spend in each state in the
long run. If not uniform, that is a bias imposed by the operator
on the system."""
import ergodic
ss = np.array(ergodic.steady_state(np.matrix(tp)))
ss = np.real(ss) # discard zero imaginary parts
ss = ss.T[0] # not sure why it ends up with an extra unused dimension
return ss
def get_steady_state(tp):
"""Given a transition probability matrix, use ergodic.steady_state
to calculate the long-run steady-state, which is a vector
representing how long the system will spend in each state in the
long run. If not uniform, that is a bias imposed by the operator
on the system."""
import ergodic
ss = np.array(ergodic.steady_state(np.matrix(tp)))
ss = np.real(ss) # discard zero imaginary parts
ss = ss.T[0] # not sure why it ends up with an extra unused dimension
return ss
def steady_state(self):
if not hasattr(self, "_steady_state"):
self._steady_state = steady_state(self.p, fill_empty_classes=True)
return self._steady_state
def s(self):
if not hasattr(self, "_s"):
self._s = steady_state(self.p)
return self._s