def sample_predictive_parameters(self):
Lext = \
np.vstack((self.L, np.sqrt(self.sigma) * np.random.randn(1, self.dim)))
D = -((Lext[:,None,:] - Lext[None,:,:])**2).sum(2)
D += self.mu_0
D += self.mu_self * np.eye(self.N+1)
P = logistic(D)
Prow = P[-1,:]
Pcol = P[:,-1]
return Prow, Pcol
def sample_predictive_parameters(self):
Lext = \
np.vstack((self.L, np.sqrt(self.sigma) * np.random.randn(1, self.dim)))
D = -((Lext[:, None, :] - Lext[None, :, :])**2).sum(2)
D += self.mu_0
D += self.mu_self * np.eye(self.N + 1)
P = logistic(D)
Prow = P[-1, :]
Pcol = P[:, -1]
return Prow, Pcol
def P(self):
P = logistic(self.D)
return P
def P(self):
P = logistic(self.D)
return P
