POSEIDON.retrieval

Functions related to atmospheric retrieval.

Attributes

`comm`
`rank`
`allowed_simplex`
`allowed_simplex_surfaces`

Functions

`run_retrieval`(planet, star, model, opac, data, priors, ...)	ADD DOCSTRING (will hopefully be done before the heat death of the Universe)
`forward_model`(param_vector, planet, star, model, opac, ...)	ADD DOCSTRING
`CLR_Prior`(chem_params_drawn[, limit])	Implements the centred-log-ratio (CLR) prior for chemical mixing ratios.
`PyMultiNest_retrieval`(planet, star, model, opac, data, ...)	Main function for conducting atmospheric retrievals with PyMultiNest.
`retrieved_samples`(planet, star, model, opac, data, ...)	ADD DOCSTRING
`get_retrieved_atmosphere`(planet, model, P[, ...])	Creates the atmosphere dictionary for the median or best fit spectrum of a retrieval.
`Z_to_sigma`(ln_Z1, ln_Z2)	Convert the log-evidences of two models to a sigma confidence level.
`Bayesian_model_comparison`(planet_name, model_1, model_2)	Conduct Bayesian model comparison between the outputs of two retrievals.

Module Contents

POSEIDON.retrieval.comm

POSEIDON.retrieval.rank

POSEIDON.retrieval.allowed_simplex = 1

POSEIDON.retrieval.allowed_simplex_surfaces = 1

POSEIDON.retrieval.run_retrieval(planet, star, model, opac, data, priors, wl, P, P_ref=None, R_p_ref=None, P_param_set=0.01, R=None, retrieval_name=None, He_fraction=0.17, N_slice_EM=2, N_slice_DN=4, constant_gravity=False, spectrum_type='transmission', y_p=np.array([0.0]), stellar_T_step=20, stellar_log_g_step=0.1, N_live=400, ev_tol=0.5, sampling_algorithm='MultiNest', resume=False, verbose=True, sampling_target='parameter', chem_grid='fastchem', N_output_samples=1000, save_ymodel=False): ADD DOCSTRING (will hopefully be done before the heat death of the Universe)

POSEIDON.retrieval.forward_model(param_vector, planet, star, model, opac, data, wl, P, P_ref_set, R_p_ref_set, P_param_set, He_fraction, N_slice_EM, N_slice_DN, spectrum_type, T_phot_grid, T_het_grid, log_g_phot_grid, log_g_het_grid, I_phot_grid, I_het_grid, y_p, F_s_obs, constant_gravity, chemistry_grid): ADD DOCSTRING

POSEIDON.retrieval.CLR_Prior(chem_params_drawn, limit=-12.0)

Implements the centred-log-ratio (CLR) prior for chemical mixing ratios.

CLR[i] here is the centred log-ratio transform of the mixing ratio, X[i]

POSEIDON.retrieval.PyMultiNest_retrieval(planet, star, model, opac, data, prior_types, prior_ranges, spectrum_type, wl, P, P_ref_set, R_p_ref_set, P_param_set, He_fraction, N_slice_EM, N_slice_DN, N_params, T_phot_grid, T_het_grid, log_g_phot_grid, log_g_het_grid, I_phot_grid, I_het_grid, y_p, F_s_obs, constant_gravity, chemistry_grid, **kwargs): Main function for conducting atmospheric retrievals with PyMultiNest.

POSEIDON.retrieval.retrieved_samples(planet, star, model, opac, data, retrieval_name, wl, P, P_ref_set, R_p_ref_set, P_param_set, He_fraction, N_slice_EM, N_slice_DN, spectrum_type, T_phot_grid, T_het_grid, log_g_phot_grid, log_g_het_grid, I_phot_grid, I_het_grid, y_p, F_s_obs, constant_gravity, chemistry_grid, N_output_samples, comm=None): ADD DOCSTRING

POSEIDON.retrieval.get_retrieved_atmosphere(planet, model, P, P_ref_set=10, R_p_ref_set=None, median=False, best_fit=True, P_param_set=0.01, He_fraction=0.17, N_slice_EM=2, N_slice_DN=4, constant_gravity=False, chemistry_grid=None, specific_param_values=[], verbose=False)

Creates the atmosphere dictionary for the median or best fit spectrum of a retrieval.

Parameters:

planet (dict) – Collection of planetary properties used POSEIDON.
model (dict) – A specific description of a given POSEIDON model.
P (np.array of float) – Model pressure grid (bar).
P_ref (float) – Reference pressure (bar).
R_p_ref (float) – Planet radius corresponding to reference pressure (m).
median (optional, bool) – Option to create an atmosphere from the median retrieved spectrum.
best_fit (optional, bool) –

Option to create an atmosphere from the best fit retrieved spectrum
He_fraction (float):

Assumed H2/He ratio (0.17 default corresponds to the solar ratio).
N_slice_EM (even int) – Number of azimuthal slices in the evening-morning transition region.
N_slice_DN (even int) – Number of zenith slices in the day-night transition region.
constant_gravity (bool) – If True, disable inverse square law gravity (only for testing).
chemistry_grid (dict) – For models with a pre-computed chemistry grid only, this dictionary is produced in chemistry.py.
specific_param_values (list) – If a specific parameter combination is provided, this will be used instead of the median or best fit parameters.

Returns:

Collection of atmospheric properties required to compute the resultant spectrum of the planet.

Return type:

atmosphere (dict)

POSEIDON.retrieval.Z_to_sigma(ln_Z1, ln_Z2): Convert the log-evidences of two models to a sigma confidence level.

POSEIDON.retrieval.Bayesian_model_comparison(planet_name, model_1, model_2, ln_Z_format='{:.2f}', B_format='{:.2e}', ln_B_format='{:.2f}', sigma_format='{:.1f}'): Conduct Bayesian model comparison between the outputs of two retrievals. This function outputs the Bayes factor and equivalent sigma significance comparing the two models.