Open phajy opened 2 months ago
Need to do some debugging and check the priors as something has gone wrong. Parameter values are unreasonably well constrained. Could this be related to cacheing - does the fit result not change if the parameters values are changed only subtly?
Iterations = 1001:1:6000
Number of chains = 1
Samples per chain = 5000
Wall duration = 3013.45 seconds
Compute duration = 3013.45 seconds
parameters = K_1, a_1, θ_1, h_1, η_1, K_2, a_2, K_3
internals = lp, n_steps, is_accept, acceptance_rate, log_density, hamiltonian_energy, hamiltonian_energy_error, max_hamiltonian_energy_error, tree_depth, numerical_error, step_size, nom_step_size
Summary Statistics
parameters mean std mcse ess_bulk ess_tail rhat ess_per_sec
Symbol Float64 Float64 Float64 Float64 Float64 Float64 Float64
K_1 0.6016 0.0000 0.0000 10.5620 18.3650 2.1226 0.0035
a_1 0.7657 0.0000 0.0000 12.9703 19.2807 1.1503 0.0043
θ_1 44.6480 0.0000 0.0000 12.2089 28.0058 1.5379 0.0041
h_1 2.2184 0.0000 0.0000 12.6782 24.9203 1.2570 0.0042
η_1 0.8199 0.0000 0.0000 11.0593 26.5691 2.0740 0.0037
K_2 0.6686 0.0000 0.0000 11.1157 18.4005 2.0280 0.0037
a_2 2.1481 0.0000 0.0000 11.0868 23.5977 1.7374 0.0037
K_3 0.3349 0.0000 0.0000 10.5624 18.3610 2.1230 0.0035
Quantiles
parameters 2.5% 25.0% 50.0% 75.0% 97.5%
Symbol Float64 Float64 Float64 Float64 Float64
K_1 0.6016 0.6016 0.6016 0.6016 0.6016
a_1 0.7657 0.7657 0.7657 0.7657 0.7657
θ_1 44.6480 44.6480 44.6480 44.6480 44.6480
h_1 2.2184 2.2184 2.2184 2.2184 2.2184
η_1 0.8199 0.8199 0.8199 0.8199 0.8199
K_2 0.6686 0.6686 0.6686 0.6686 0.6686
a_2 2.1481 2.1481 2.1481 2.1481 2.1481
K_3 0.3349 0.3349 0.3349 0.3349 0.3349
Might skip this for the talk. Can add later, and will need for the paper.
WIP to perform MCMC fitting, in this case for a thick disc model. I've not run this yet, but it would benefit from parallelisation.