def update_zeta_j_wrapper(args):
# parse arguments
curr_zeta_j, n_j, Y_js, lY_js, alpha, beta, xi, tau, sigma_unity = args
prop_zeta_j = np.empty(curr_zeta_j.shape)
# prepare placeholders for xi, tau (with same indexing as zeta)
xxi = np.ones(curr_zeta_j.shape)
xta = np.ones(curr_zeta_j.shape)
xxi[~sigma_unity] = xi
xta[~sigma_unity] = tau
# iterate through zeta sampling
for i in range(curr_zeta_j.shape[0]):
if sigma_unity[i]:
prop_zeta_j[i] = sample_alpha_1_mh_summary(
curr_zeta_j[i],
n_j,
Y_js[i],
lY_js[i],
alpha[i],
beta[i],
)
else:
prop_zeta_j[i] = sample_alpha_k_mh_summary(
curr_zeta_j[i],
n_j,
Y_js[i],
lY_js[i],
alpha[i],
beta[i],
xxi[i],
xta[i],
)
return prop_zeta_j
def update_zeta_j_wrapper(args):
# parse arguments
curr_zeta_j, n_j, Y_js, lY_js, alpha, beta, xi, tau = args
prop_zeta_j = np.empty(curr_zeta_j.shape)
prop_zeta_j[0] = sample_alpha_1_mh_summary(
curr_zeta_j[0], n_j, Y_js[0], lY_js[0], alpha[0], beta[0]
)
for i in range(1, curr_zeta_j.shape[0]):
prop_zeta_j[i] = sample_alpha_k_mh_summary(
curr_zeta_j[i], n_j, Y_js[i], lY_js[i],
alpha[i], beta[i], xi[i-1], tau[i-1],
)
return prop_zeta_j
def update_zeta_wrapper(args):
# parse arguments
curr_zeta, n, Ys, lYs, alpha, beta, xi, tau = args
prop_zeta = np.empty(curr_zeta.shape)
prop_zeta[0] = sample_alpha_1_mh_summary(
curr_zeta[0], n, Ys[0], lYs[0], alpha, beta
)
for i in range(1, curr_zeta.shape[0]):
prop_zeta[i] = sample_alpha_k_mh_summary(
curr_zeta[i], n, Ys[i], lYs[i],
alpha, beta, xi, tau,
)
return prop_zeta
def sample_gamma_shape_wrapper(args):
return sample_alpha_k_mh_summary(*args)
def sample_xi_wrapper(args):
return sample_alpha_k_mh_summary(*args)