petitRADTRANS.retrieval.models ============================== .. py:module:: petitRADTRANS.retrieval.models .. autoapi-nested-parse:: Models Module This module contains a set of functions that generate the spectra used in the petitRADTRANS retrieval. This includes setting up the pressure-temperature structure, the chemistry, and the radiative transfer to compute the emission or transmission spectrum. See the documentation on the Retrieval Models for full details. All models must take the same set of inputs: ModelContext : ModelContext An object that contains the static information about the retrieval, including the Radtrans object. PhysicalParams : PhysicalParams An object containing all of the physical parameters of the model, which can be free parameters, or fixed values. pt_plot_mode : bool If this argument is True, the model function will return the pressure and temperature arrays instead of computing the flux. Attributes ---------- .. autoapisummary:: petitRADTRANS.retrieval.models.MODEL_CONTRACT_LEGACY petitRADTRANS.retrieval.models.MODEL_CONTRACT_DIFFERENTIABLE petitRADTRANS.retrieval.models.MODEL_CONTRACT_LEGACY_ADAPTER petitRADTRANS.retrieval.models._MODEL_CONTRACT_ATTRIBUTE petitRADTRANS.retrieval.models._WRAPPED_LEGACY_MODEL_ATTRIBUTE petitRADTRANS.retrieval.models._TIME_INDEXED_PATTERN petitRADTRANS.retrieval.models._SINUSOIDAL_SUFFIXES petitRADTRANS.retrieval.models._EMISSION_RUNTIME_AUXILIARY_OUTPUT_EXCLUSIONS petitRADTRANS.retrieval.models.molliere_2020_emission petitRADTRANS.retrieval.models.molliere_2020_two_column petitRADTRANS.retrieval.models.guillot_emission petitRADTRANS.retrieval.models.guillot_emission_add_gaussian_temperature petitRADTRANS.retrieval.models.guillot_patchy_emission petitRADTRANS.retrieval.models.interpolated_profile_emission petitRADTRANS.retrieval.models.gradient_profile_emission petitRADTRANS.retrieval.models.power_law_profile_emission petitRADTRANS.retrieval.models.guillot_transmission petitRADTRANS.retrieval.models.guillot_patchy_transmission petitRADTRANS.retrieval.models.madhushudhan_seager_emission petitRADTRANS.retrieval.models.madhushudhan_seager_transmission petitRADTRANS.retrieval.models.madhu_seager_patchy_transmission petitRADTRANS.retrieval.models.isothermal_transmission petitRADTRANS.retrieval.models.power_law_profile_transmission petitRADTRANS.retrieval.models.time_series_gradient_emission Functions --------- .. autoapisummary:: petitRADTRANS.retrieval.models._is_full_easychem_requested petitRADTRANS.retrieval.models._as_structural_int petitRADTRANS.retrieval.models._normalize_legacy_output_value petitRADTRANS.retrieval.models._compute_gravity petitRADTRANS.retrieval.models._resolve_guillot_gamma petitRADTRANS.retrieval.models._clip_easychem_temperatures petitRADTRANS.retrieval.models._wrap_runtime_temperature_builder petitRADTRANS.retrieval.models._stack_runtime_cloud_base_pressures petitRADTRANS.retrieval.models.molliere_2020_emission petitRADTRANS.retrieval.models.molliere_2020_two_column petitRADTRANS.retrieval.models.guillot_emission petitRADTRANS.retrieval.models.guillot_relative_emission petitRADTRANS.retrieval.models.guillot_emission_add_gaussian_temperature petitRADTRANS.retrieval.models.guillot_patchy_emission petitRADTRANS.retrieval.models.interpolated_profile_emission petitRADTRANS.retrieval.models.gradient_profile_emission petitRADTRANS.retrieval.models._build_timestep_params petitRADTRANS.retrieval.models.time_series_gradient_emission petitRADTRANS.retrieval.models.power_law_profile_emission petitRADTRANS.retrieval.models.guillot_transmission petitRADTRANS.retrieval.models.guillot_patchy_transmission petitRADTRANS.retrieval.models.madhushudhan_seager_emission petitRADTRANS.retrieval.models.madhushudhan_seager_transmission petitRADTRANS.retrieval.models.madhu_seager_patchy_transmission petitRADTRANS.retrieval.models.isothermal_transmission petitRADTRANS.retrieval.models.power_law_profile_transmission petitRADTRANS.retrieval.models.mark_legacy_model_function petitRADTRANS.retrieval.models.legacy_molliere_2020_emission petitRADTRANS.retrieval.models.legacy_molliere_2020_two_column petitRADTRANS.retrieval.models.legacy_guillot_emission petitRADTRANS.retrieval.models.legacy_guillot_emission_add_gaussian_temperature petitRADTRANS.retrieval.models.legacy_guillot_patchy_emission petitRADTRANS.retrieval.models.legacy_interpolated_profile_emission petitRADTRANS.retrieval.models.legacy_gradient_profile_emission petitRADTRANS.retrieval.models.legacy_power_law_profile_emission petitRADTRANS.retrieval.models.legacy_guillot_transmission petitRADTRANS.retrieval.models.legacy_guillot_patchy_transmission petitRADTRANS.retrieval.models.legacy_madhushudhan_seager_emission petitRADTRANS.retrieval.models.legacy_madhushudhan_seager_transmission petitRADTRANS.retrieval.models.legacy_madhu_seager_patchy_transmission petitRADTRANS.retrieval.models.legacy_isothermal_transmission petitRADTRANS.retrieval.models.legacy_power_law_profile_transmission petitRADTRANS.retrieval.models.add_blackbody_cpd_model petitRADTRANS.retrieval.models.initialize_pressure petitRADTRANS.retrieval.models.initialize_pressure_from_legacy_dict petitRADTRANS.retrieval.models._as_auxiliary_outputs petitRADTRANS.retrieval.models._as_filtered_auxiliary_outputs petitRADTRANS.retrieval.models._is_scalar_like petitRADTRANS.retrieval.models._build_model_context_from_parameters petitRADTRANS.retrieval.models._normalize_model_output petitRADTRANS.retrieval.models._model_result_to_legacy_output petitRADTRANS.retrieval.models.run_legacy_model_adapter petitRADTRANS.retrieval.models._infer_model_function_contract petitRADTRANS.retrieval.models.coerce_model_generating_function petitRADTRANS.retrieval.models.run_runtime_model_adapter petitRADTRANS.retrieval.models.make_unified_model petitRADTRANS.retrieval.models._initialize_runtime_pressure_grid petitRADTRANS.retrieval.models._resolve_runtime_abundance_state petitRADTRANS.retrieval.models._build_runtime_emission_model_inputs petitRADTRANS.retrieval.models._build_runtime_transmission_model_inputs petitRADTRANS.retrieval.models.calculate_emission_spectrum_runtime petitRADTRANS.retrieval.models.calculate_emission_spectra_batched petitRADTRANS.retrieval.models.calculate_transmission_spectrum_runtime petitRADTRANS.retrieval.models.calculate_emission_spectrum petitRADTRANS.retrieval.models.calculate_transmission_spectrum Module Contents --------------- .. py:function:: _is_full_easychem_requested(parameters) -> bool Return whether the retrieval explicitly requested the full easyChem path. .. py:data:: MODEL_CONTRACT_LEGACY :value: 'legacy' .. py:data:: MODEL_CONTRACT_DIFFERENTIABLE :value: 'differentiable' .. py:data:: MODEL_CONTRACT_LEGACY_ADAPTER :value: 'runtime_legacy_adapter' .. py:data:: _MODEL_CONTRACT_ATTRIBUTE :value: '_prt_model_contract' .. py:data:: _WRAPPED_LEGACY_MODEL_ATTRIBUTE :value: '_prt_wrapped_legacy_model' .. py:data:: _TIME_INDEXED_PATTERN .. py:data:: _SINUSOIDAL_SUFFIXES :value: ('_amplitude', '_period', '_phase', '_offset') .. py:function:: _as_structural_int(value, parameter_name: str) -> int Return a Python int for parameters that define model structure. Counts such as ``nnodes``, ``N_layers``, and ``N_time`` are used in Python control flow and array-shape construction, so they must remain compile-time constants under JAX transforms. .. py:function:: _normalize_legacy_output_value(value) .. py:function:: _compute_gravity(physical_params) .. py:function:: _resolve_guillot_gamma(parameters) Return the Guillot gamma parameter in linear units for any model contract. .. py:function:: _clip_easychem_temperatures(physical_params, temperatures) .. py:function:: _wrap_runtime_temperature_builder(physical_params, temperature_builder) .. py:function:: _stack_runtime_cloud_base_pressures(p_bases, dtype) .. py:function:: molliere_2020_emission(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: molliere_2020_two_column(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: guillot_emission(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: guillot_relative_emission(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult Guillot emission model normalised by a PHOENIX stellar spectrum. Computes the same Guillot global temperature profile and emission spectrum as :func:`guillot_emission`, then divides the planet flux by a PHOENIX stellar spectrum binned to the same wavelength grid. The returned spectrum is the dimensionless planet-to-star flux ratio F_planet / F_star, suitable for secondary-eclipse or direct-imaging contrast retrieval. Required parameters (shared with :func:`guillot_emission`) ----------------------------------------------------------- distance_to_system : float Distance from the system to the observer in cm. log_g, planet_radius, mass : float Two of these three must be supplied (see :func:`_compute_gravity`). T_int, T_equ, gamma, log_kappa_IR : float Guillot temperature-profile parameters. Fe/H, C/O, log_pquench : float Equilibrium-chemistry parameters. Additional stellar parameters ----------------------------- star_effective_temperature : float Effective temperature of the host star in K. Used to interpolate the PHOENIX spectral grid. stellar_radius : float, optional Stellar radius in cm. If omitted, the radius from the PHOENIX grid is used. .. py:function:: guillot_emission_add_gaussian_temperature(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: guillot_patchy_emission(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: interpolated_profile_emission(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: gradient_profile_emission(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: _build_timestep_params(physical_params, t_idx, t_value) Build a :class:`PhysicalParams` snapshot for a single model timestep. Auto-detects time-varying parameters by scanning parameter names: - **Per-timestep keys** matching ``{P}_t_{int}``: the value at index *t_idx* is stored under the base name ``P``. - **Sinusoidal quartets** ``{P}_amplitude``, ``{P}_period``, ``{P}_phase``, ``{P}_offset``: the base name ``P`` is evaluated as ``offset + amplitude * sin(2π * t_value / period + phase)``. All detected meta-keys are stripped from the snapshot; only the resolved base name is included. Parameters matching neither pattern are copied unchanged. This allows *any* parameter to vary with time simply by following the naming convention—no pre-declaration is required. .. py:function:: time_series_gradient_emission(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult Time-variable gradient-profile emission model. Computes ``N_time`` emission spectra at regularly spaced model timesteps spanning the range of the observation times. Spectra at intermediate observation times are obtained by linear interpolation along the time axis. Time-varying parameters are **auto-detected** from the naming convention by :func:`_build_timestep_params` — no explicit declaration is needed. Any parameter can vary with time simply by following one of the two supported patterns: Per-timestep keys ----------------- Supply ``{P}_t_0`` … ``{P}_t_{N_time-1}`` as free parameters. At each model timestep *i* the snapshot will contain ``P`` set to the value of ``{P}_t_i``. Sinusoidal parameterisation --------------------------- Supply ``{P}_amplitude``, ``{P}_period`` (seconds), ``{P}_phase`` (radians), and ``{P}_offset``. The value at time *t* is .. math:: P(t) = \text{offset} + A \sin\!\left(\frac{2\pi t}{\text{period}} + \phi\right) Parameters that do not match either pattern are treated as constant across all timesteps (e.g. ``distance_to_system``, ``mass``, ``planet_radius``, ``N_layers``). Returns ------- ModelResult When ``pt_plot_mode=False``: ``spectrum`` has shape ``(N_time, N_wavelength)`` on the shared model time grid, and ``auxiliary_outputs['model_times']`` stores that grid. The runtime scorer projects each epoch to observation space before interpolating in time, so the model only needs to carry one spectral representation. When ``pt_plot_mode=True``: ``temperatures`` has shape ``(N_time, N_pressure)`` and ``auxiliary_outputs['model_times']`` gives the corresponding model time grid. .. py:function:: power_law_profile_emission(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: guillot_transmission(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: guillot_patchy_transmission(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: madhushudhan_seager_emission(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: madhushudhan_seager_transmission(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: madhu_seager_patchy_transmission(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: isothermal_transmission(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: power_law_profile_transmission(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: mark_legacy_model_function(model_function) .. py:function:: legacy_molliere_2020_emission(prt_object, parameters, pt_plot_mode=False, adaptive_mesh_refinement=True) Disequilibrium Chemistry Emission Model This model computes an emission spectrum based on the temperature profile of (Molliere 2020). (Dis)equilibrium or free chemistry, can be used. The use of easychem for on-the-fly (dis)equilibrium chemistry calculations is supported, but is currently under development. Many of the parameters are optional, but must be used in the correct combination with other parameters. Args: prt_object : object An instance of the pRT class, with optical properties as defined in the RunDefinition. parameters : dict Dictionary of required parameters: * distance_to_system : Distance to the planet in [cm] Two of * log_g : Log of surface gravity * planet_radius : planet radius [cm] * mass : planet mass [g] * T_int : Interior temperature of the planet [K] * T3 : Innermost temperature spline [K] * T2 : Middle temperature spline [K] * T1 : Outer temperature spline [K] * alpha : power law index in tau = delta * press_cgs**alpha * log_delta : proportionality factor in tau = delta * press_cgs**alpha * log_pquench : Pressure at which CO, CH4 and H2O abundances become vertically constant * Fe/H : Metallicity * C/O : Carbon to oxygen ratio * fsed : sedimentation parameter - can be unique to each cloud type by adding _CloudName One of: * sigma_lnorm : Width of cloud particle size distribution (log normal) * b_hans : Width of cloud particle size distribution (hansen) One of: * log_cloud_radius_* : Central particle radius (typically computed with fsed and Kzz) * log_kzz : Vertical mixing parameter One of * eq_scaling_* : Scaling factor for equilibrium cloud abundances. * log_X_cb_: cloud mass fraction abundance Optional * contribution : return the emission contribution function * patchiness : Cloud coverage fraction, mixes two columns with different cloud properties. * remove_cloud_species : Specifies which cloud species to remove for clear atmosphere column. * T_disk_blackbody : Temperature of a blackbody circumplanetary disk component. * disk_radius : Radius [cm] of a blackbody circumplanetary disk component. pt_plot_mode : bool Return only the pressure-temperature profile for plotting. Evaluate mode only. adaptive_mesh_refinement : bool Adaptive mesh refinement. Use the high resolution pressure grid around the cloud base. Returns: wlen_model : ArrayLike Wavlength array of computed model, not binned to data [um] spectrum_model : ArrayLike Computed emission spectrum [W/m2/micron] contr_em : Optional, ArrayLike Emission contribution function, relative contributions for each wavelength and pressure level. .. py:function:: legacy_molliere_2020_two_column(prt_object, parameters, pt_plot_mode=False, adaptive_mesh_refinement=True) Disequilibrium Chemistry Emission Model This model computes an emission spectrum based on the temperature profile of (Molliere 2020). (Dis)equilibrium or free chemistry, can be used. The use of easychem for on-the-fly (dis)equilibrium chemistry calculations is supported, but is currently under development. This model includes patchy clouds, and requires a unique temperature profile for the clear atmosphere regions - ie this is a full two column model! Args: prt_object : object An instance of the pRT class, with optical properties as defined in the RunDefinition. parameters : dict Dictionary of required parameters: * distance_to_system : Distance to the planet in [cm] Two of * log_g : Log of surface gravity * planet_radius : planet radius [cm] * mass : planet mass [g] * T_int : Interior temperature of the planet [K] * T3 : Innermost temperature spline * T2 : Middle temperature spline * T1 : Outer temperature spline * T3_clear : Innermost temperature spline for clear atmosphere * T2_clear : Middle temperature spline for clear atmosphere * T1_clear : Outer temperature spline for clear atmosphere * alpha : power law index in tau = delta * press_cgs**alpha * alpha_clear : power law index in tau = delta * press_cgs**alpha for clear atmosphere * log_delta : proportionality factor in tau = delta * press_cgs**alpha * log_delta_clear : proportionality factor in tau = delta * press_cgs**alpha for clear atmosphere * log_pquench : Pressure at which CO, CH4 and H2O abundances become vertically constant * Fe/H : Metallicity * C/O : Carbon to oxygen ratio * fsed : sedimentation parameter - can be unique to each cloud type One of: * sigma_lnorm : Width of cloud particle size distribution (log normal) * b_hans : Width of cloud particle size distribution (hansen) One of: * log_cloud_radius_* : Central particle radius (typically computed with fsed and Kzz) * log_kzz : Vertical mixing parameter One of * eq_scaling_* : Scaling factor for equilibrium cloud abundances. * log_X_cb_: cloud mass fraction abundance Optional * contribution : return the emission contribution function * patchiness : Cloud coverage fraction, mixes two columns with different cloud properties. * remove_cloud_species : Specifies which cloud species to remove for the clear atmosphere column. * T_disk_blackbody : Temperature of a blackbody circumplanetary disk component. * disk_radius : Radius [cm] of a blackbody circumplanetary disk component. pt_plot_mode : bool Return only the pressure-temperature profile for plotting. Evaluate mode only. adaptive_mesh_refinement : bool Adaptive mesh refinement. Use the high resolution pressure grid around the cloud base. Returns: wlen_model : jnp.array Wavlength array of computed model, not binned to data [um] spectrum_model : jnp.array Computed emission spectrum [W/m2/micron] contr_em : Optional, ArrayLike Emission contribution function, relative contributions for each wavelength and pressure level. Only returns the contribution of the clear atmosphere component. .. py:function:: legacy_guillot_emission(prt_object, parameters, pt_plot_mode=False, adaptive_mesh_refinement=False) Emission spectrum calculation for the Guillot 2010 temperature profile. (Dis)equilibrium or free chemistry, can be used. The use of easychem for on-the-fly (dis)equilibrium chemistry calculations is supported, but is currently under development. Args: prt_object : object An instance of the pRT class, with optical properties as defined in the RunDefinition. parameters : dict Dictionary of required parameters: * distance_to_system : Distance to the planet in [cm] Two of * log_g : Log of surface gravity * planet_radius : planet radius [cm] * mass : planet mass [g] * T_int : Interior temperature of the planet [K] * T_equ : Equilibrium temperature of the planet * gamma : Guillot gamma parameter * log_kappa_IR : The log of the ratio between the infrared and optical opacities Either: * log_pquench : Pressure at which CO, CH4 and H2O abundances become vertically constant * Fe/H : Metallicity * C/O : Carbon to oxygen ratio Or: * $SPECIESNAME[_$DATABASE][_R_$RESOLUTION] : The log mass fraction abundance of the species * fsed : sedimentation parameter - can be unique to each cloud type One of: * sigma_lnorm : Width of cloud particle size distribution (log normal) * b_hans : Width of cloud particle size distribution (hansen) One of: * log_cloud_radius_* : Central particle radius (typically computed with fsed and Kzz) * log_kzz : Vertical mixing parameter One of * eq_scaling_* : Scaling factor for equilibrium cloud abundances. * log_X_cb_: cloud mass fraction abundance Optional * contribution : return the emission contribution function * patchiness : Cloud coverage fraction, mixes two columns with different cloud properties. * remove_cloud_species : Specifies which cloud species to remove for the clear atmosphere column. * T_disk_blackbody : Temperature of a blackbody circumplanetary disk component. * disk_radius : Radius [cm] of a blackbody circumplanetary disk component. pt_plot_mode : bool Return only the pressure-temperature profile for plotting. Evaluate mode only. adaptive_mesh_refinement : Adaptive mesh refinement. Use the high resolution pressure grid around the cloud base. Returns: wlen_model : jnp.array Wavlength array of computed model, not binned to data [um] spectrum_model : jnp.array Computed transmission spectrum planet_radius**2/Rstar**2 contr-em : ArrayLike Optional, Emission contribution function, relative contributions for each wavelength and pressure level. .. py:function:: legacy_guillot_emission_add_gaussian_temperature(prt_object, parameters, pt_plot_mode=False, adaptive_mesh_refinement=False) Emission spectrum calculation for the Guillot 2010 temperature profile. (Dis)equilibrium or free chemistry, can be used. The use of easychem for on-the-fly (dis)equilibrium chemistry calculations is supported, but is currently under development. Args: prt_object : object An instance of the pRT class, with optical properties as defined in the RunDefinition. parameters : dict Dictionary of required parameters: * distance_to_system : Distance to the planet in [cm] Two of * log_g : Log of surface gravity * planet_radius : planet radius [cm] * mass : planet mass [g] * T_int : Interior temperature of the planet [K] * T_equ : Equilibrium temperature of the planet * gamma : Guillot gamma parameter * log_kappa_IR : The log of the ratio between the infrared and optical opacities Either: * log_pquench : Pressure at which CO, CH4 and H2O abundances become vertically constant * Fe/H : Metallicity * C/O : Carbon to oxygen ratio Or: * $SPECIESNAME[_$DATABASE][_R_$RESOLUTION] : The log mass fraction abundance of the species * fsed : sedimentation parameter - can be unique to each cloud type One of: * sigma_lnorm : Width of cloud particle size distribution (log normal) * b_hans : Width of cloud particle size distribution (hansen) One of: * log_cloud_radius_* : Central particle radius (typically computed with fsed and Kzz) * log_kzz : Vertical mixing parameter One of * eq_scaling_* : Scaling factor for equilibrium cloud abundances. * log_X_cb_: cloud mass fraction abundance Optional * contribution : return the emission contribution function * patchiness : Cloud coverage fraction, mixes two columns with different cloud properties. * remove_cloud_species : Specifies which cloud species to remove for the clear atmosphere column. * T_disk_blackbody : Temperature of a blackbody circumplanetary disk component. * disk_radius : Radius [cm] of a blackbody circumplanetary disk component. pt_plot_mode : bool Return only the pressure-temperature profile for plotting. Evaluate mode only. adaptive_mesh_refinement : Adaptive mesh refinement. Use the high resolution pressure grid around the cloud base. Returns: wlen_model : jnp.array Wavlength array of computed model, not binned to data [um] spectrum_model : jnp.array Computed transmission spectrum planet_radius**2/Rstar**2 contr-em : ArrayLike Optional, Emission contribution function, relative contributions for each wavelength and pressure level. .. py:function:: legacy_guillot_patchy_emission(prt_object, parameters, pt_plot_mode=False, adaptive_mesh_refinement=False) Deprecated, to be removed in future version .. py:function:: legacy_interpolated_profile_emission(prt_object, parameters, pt_plot_mode=False, adaptive_mesh_refinement=False) This model computes a emission spectrum based a spline temperature-pressure profile. Either free or equilibrium chemistry can be used, together with a range of cloud parameterizations. It is possible to use free abundances for some species and equilibrium chemistry for the remainder. Args: prt_object : object An instance of the pRT class, with optical properties as defined in the RunDefinition. parameters : dict Dictionary of required parameters: * distance_to_system : Distance to the planet in [cm] Two of * log_g : Log of surface gravity * planet_radius : planet radius [cm] * mass : planet mass [g] * nnodes : number of nodes to interplate, excluding the first and last points. so the total number of nodes is nnodes + 2 * T{i} : One parameter for each temperature node * gamma : weight for penalizing the profile curvature Either: * log_pquench : Pressure at which CO, CH4 and H2O abundances become vertically constant * Fe/H : Metallicity * C/O : Carbon to oxygen ratio Or: * $SPECIESNAME[_$DATABASE][_R_$RESOLUTION] : The log mass fraction abundance of the species Optional: * fsed : sedimentation parameter - can be unique to each cloud type One of: * sigma_lnorm : Width of cloud particle size distribution (log normal) * b_hans : Width of cloud particle size distribution (hansen) One of: * log_cloud_radius_* : Central particle radius (typically computed with fsed and Kzz) * log_kzz : Vertical mixing parameter One of * eq_scaling_* : Scaling factor for equilibrium cloud abundances. * log_X_cb_: cloud mass fraction abundance Optional * contribution : return the emission contribution function * patchiness : Cloud coverage fraction, mixes two columns with different cloud properties. * remove_cloud_species : Specifies which cloud species to remove for the clear atmosphere column. * T_disk_blackbody : Temperature of a blackbody circumplanetary disk component. * disk_radius : Radius [cm] of a blackbody circumplanetary disk component. pt_plot_mode : bool Return only the pressure-temperature profile for plotting. Evaluate mode only. adaptive_mesh_refinement : Adaptive mesh refinement. Use the high resolution pressure grid around the cloud base. Returns: wlen_model : jnp.array Wavlength array of computed model, not binned to data [um] spectrum_model : jnp.array Computed transmission spectrum planet_radius**2/Rstar**2 contr-em : ArrayLike Optional, the emission contribution function, relative contributions for each wavelength and pressure level. .. py:function:: legacy_gradient_profile_emission(prt_object, parameters, pt_plot_mode=False, adaptive_mesh_refinement=False) This model computes a emission spectrum based a gradient temperature-pressure profile (Zhang 2023). Either free or equilibrium chemistry can be used, together with a range of cloud parameterizations. It is possible to use free abundances for some species and equilibrium chemistry for the remainder. Args: prt_object : object An instance of the pRT class, with optical properties as defined in the RunDefinition. parameters : dict Dictionary of required parameters: * distance_to_system : Distance to the planet in [cm] Two of * log_g : Log of surface gravity * planet_radius : planet radius [cm] * mass : planet mass [g] * N_layers : number of nodes to interplate, excluding the first and last points. so the total number of nodes is nnodes + 2 * T_bottom : temperature at the base of the atmosphere * PTslope_* : temperature gradient for each of the n_layers between which the profile is interpolated. Either: * log_pquench : Pressure at which CO, CH4 and H2O abundances become vertically constant * Fe/H : Metallicity * C/O : Carbon to oxygen ratio Or: * $SPECIESNAME[_$DATABASE][_R_$RESOLUTION] : The log mass fraction abundance of the species Optional: * fsed : sedimentation parameter - can be unique to each cloud type One of: * sigma_lnorm : Width of cloud particle size distribution (log normal) * b_hans : Width of cloud particle size distribution (hansen) One of: * log_cloud_radius_* : Central particle radius (typically computed with fsed and Kzz) * log_kzz : Vertical mixing parameter One of * eq_scaling_* : Scaling factor for equilibrium cloud abundances. * log_X_cb_: cloud mass fraction abundance Optional * contribution : return the emission contribution function * patchiness : Cloud coverage fraction, mixes two columns with different cloud properties. * remove_cloud_species : Specifies which cloud species to remove for the clear atmosphere column. * T_disk_blackbody : Temperature of a blackbody circumplanetary disk component. * disk_radius : Radius [cm] of a blackbody circumplanetary disk component. pt_plot_mode : bool Return only the pressure-temperature profile for plotting. Evaluate mode only. adaptive_mesh_refinement : Adaptive mesh refinement. Use the high resolution pressure grid around the cloud base. Returns: wlen_model : jnp.array Wavlength array of computed model, not binned to data [um] spectrum_model : jnp.array Computed transmission spectrum planet_radius**2/Rstar**2 contr-em : ArrayLike Optional, the emission contribution function, relative contributions for each wavelength and pressure level. .. py:function:: legacy_power_law_profile_emission(prt_object, parameters, pt_plot_mode=False, adaptive_mesh_refinement=False) This model computes a emission spectrum based a gradient temperature-pressure profile (Zhang 2023). Either free or equilibrium chemistry can be used, together with a range of cloud parameterizations. It is possible to use free abundances for some species and equilibrium chemistry for the remainder. Args: prt_object : object An instance of the pRT class, with optical properties as defined in the RunDefinition. parameters : dict Dictionary of required parameters: * distance_to_system : Distance to the planet in [cm] Two of * log_g : Log of surface gravity * planet_radius : planet radius [cm] * mass : planet mass [g] * alpha : power law slope for the temperture profile * T_0 : multiplicative factor for the power law slope Either: * log_pquench : Pressure at which CO, CH4 and H2O abundances become vertically constant * Fe/H : Metallicity * C/O : Carbon to oxygen ratio Or: * $SPECIESNAME[_$DATABASE][_R_$RESOLUTION] : The log mass fraction abundance of the species Optional: * fsed : sedimentation parameter - can be unique to each cloud type One of: * sigma_lnorm : Width of cloud particle size distribution (log normal) * b_hans : Width of cloud particle size distribution (hansen) One of: * log_cloud_radius_* : Central particle radius (typically computed with fsed and Kzz) * log_kzz : Vertical mixing parameter One of * eq_scaling_* : Scaling factor for equilibrium cloud abundances. * log_X_cb_: cloud mass fraction abundance Optional * contribution : return the emission contribution function * patchiness : Cloud coverage fraction, mixes two columns with different cloud properties. * remove_cloud_species : Specifies which cloud species to remove for the clear atmosphere column. * T_disk_blackbody : Temperature of a blackbody circumplanetary disk component. * disk_radius : Radius [cm] of a blackbody circumplanetary disk component. pt_plot_mode : bool Return only the pressure-temperature profile for plotting. Evaluate mode only. adaptive_mesh_refinement : Adaptive mesh refinement. Use the high resolution pressure grid around the cloud base. Returns: wlen_model : jnp.array Wavlength array of computed model, not binned to data [um] spectrum_model : jnp.array Computed transmission spectrum planet_radius**2/Rstar**2 contr-em : ArrayLike Optional, the emission contribution function, relative contributions for each wavelength and pressure level. .. py:function:: legacy_guillot_transmission(prt_object, parameters, pt_plot_mode=False, adaptive_mesh_refinement=False) Transmission Model, Guillot Profile This model computes a transmission spectrum based on the Guillot profile Either free or (dis)equilibrium chemistry can be used, together with a range of cloud parameterizations. It is possible to use free abundances for some species and equilibrium chemistry for the remainder. Chemical clouds can be used, or a simple gray opacity source. Args: prt_object : object An instance of the pRT class, with optical properties as defined in the RunDefinition. parameters : dict Dictionary of required parameters: * distance_to_system : Distance to the planet in [cm] Two of * log_g : Log of surface gravity * planet_radius : planet radius [cm] * mass : planet mass [g] * T_int : Interior temperature of the planet [K] * T_equ : Equilibrium temperature of the planet * gamma : Guillot gamma parameter * log_kappa_IR : The log of the ratio between the infrared and optical opacities Either: * log_pquench : Pressure at which CO, CH4 and H2O abundances become vertically constant * Fe/H : Metallicity * C/O : Carbon to oxygen ratio Or: * $SPECIESNAME[_$DATABASE][.R$RESOLUTION] : The log mass fraction abundance of the species Either: * [log_]Pcloud : The (log) pressure at which to place the gray cloud opacity. Or: * fsed : sedimentation parameter - can be unique to each cloud type One of: * sigma_lnorm : Width of cloud particle size distribution (log normal) * b_hans : Width of cloud particle size distribution (hansen) One of: * log_cloud_radius_* : Central particle radius (typically computed with fsed and Kzz) * log_kzz : Vertical mixing parameter One of * eq_scaling_* : Scaling factor for equilibrium cloud abundances. * log_X_cb_: cloud mass fraction abundance Optional * contribution : return the transmission contribution function * power_law_opacity_coefficient : gamma, power law slope for a rayleigh-like haze * haze_factor : multiplicative scaling factor for the strength of the rayleigh haze * power_law_opacity_350nm : strength of the rayleigh haze at 350 nm. pt_plot_mode : bool Return only the pressure-temperature profile for plotting. Evaluate mode only. adaptive_mesh_refinement : Adaptive mesh refinement. Use the high resolution pressure grid around the cloud base. Returns: wlen_model : jnp.array Wavlength array of computed model, not binned to data [um] spectrum_model : jnp.array Computed transmission spectrum planet_radius**2/Rstar**2 contr-em : ArrayLike Optional, the transmission contribution function, relative contributions for each wavelength and pressure level. .. py:function:: legacy_guillot_patchy_transmission(prt_object, parameters, pt_plot_mode=False, adaptive_mesh_refinement=False) Deprecated .. py:function:: legacy_madhushudhan_seager_emission(prt_object, parameters, pt_plot_mode=False, adaptive_mesh_refinement=False) Transmission Model, Madhusudhan Seager 2009 Profile This model computes a transmission spectrum based on a Guillot temperature-pressure profile. Either free or equilibrium chemistry can be used, together with a range of cloud parameterizations. It is possible to use free abundances for some species and equilibrium chemistry for the remainder. Chemical clouds can be used, or a simple gray opacity source. This model requires patchy clouds. Args: prt_object : object An instance of the pRT class, with optical properties as defined in the RunDefinition. parameters : dict Dictionary of required parameters: * distance_to_system : Distance to the planet in [cm] Two of * log_g : Log of surface gravity * planet_radius : planet radius [cm] * mass : planet mass [g] * log_P_set : Pressure value to contrain the PT profile, defaults to 10 bar. * T_set : temperature at P_set to constrain the PT profile. [K] * log_P3 : (log) Pressure value for the top of the deep atmospheric layer, [bar] * P2 : in range (0,1), sets the pressure level of the middle atmospheric layer * P1 : in range (0,1), sets the pressure level of the top atmospheric layer * alpha_0 : slope of the upper atmospheric layer * alpha_1 : slope of the middle atmospheric layer Optional : * beta : power law for the slopes, default value is 0.5. Not recommended to change! Either: * log_pquench : Pressure at which CO, CH4 and H2O abundances become vertically constant * Fe/H : Metallicity * C/O : Carbon to oxygen ratio Or: * $SPECIESNAME[_$DATABASE][.R$RESOLUTION] : The log mass fraction abundance of the species Optional: * fsed : sedimentation parameter - can be unique to each cloud type One of: * sigma_lnorm : Width of cloud particle size distribution (log normal) * b_hans : Width of cloud particle size distribution (hansen) One of: * log_cloud_radius_* : Central particle radius (typically computed with fsed and Kzz) * log_kzz : Vertical mixing parameter One of * eq_scaling_* : Scaling factor for equilibrium cloud abundances. * log_X_cb_: cloud mass fraction abundance Optional * contribution : return the emission contribution function * patchiness : Cloud coverage fraction, mixes two columns with different cloud properties. * remove_cloud_species : Specifies which cloud species to remove for the clear atmosphere column. * T_disk_blackbody : Temperature of a blackbody circumplanetary disk component. * disk_radius : Radius [cm] of a blackbody circumplanetary disk component. pt_plot_mode : bool Return only the pressure-temperature profile for plotting. Evaluate mode only. adaptive_mesh_refinement : Adaptive mesh refinement. Use the high resolution pressure grid around the cloud base. Returns: wlen_model : jnp.array Wavlength array of computed model, not binned to data [um] spectrum_model : jnp.array Computed transmission spectrum planet_radius**2/Rstar**2 contr-em : ArrayLike Optional, the transmission contribution function, relative contributions for each wavelength and pressure level. Only the clear atmosphere contribution is returned. .. py:function:: legacy_madhushudhan_seager_transmission(prt_object, parameters, pt_plot_mode=False, adaptive_mesh_refinement=False) Transmission Model, Madhusudhan Seager 2009 Profile This model computes a transmission spectrum based on a Guillot temperature-pressure profile. Either free or equilibrium chemistry can be used, together with a range of cloud parameterizations. It is possible to use free abundances for some species and equilibrium chemistry for the remainder. Chemical clouds can be used, or a simple gray opacity source. This model requires patchy clouds. Args: prt_object : object An instance of the pRT class, with optical properties as defined in the RunDefinition. parameters : dict Dictionary of required parameters: * distance_to_system : Distance to the planet in [cm] Two of * log_g : Log of surface gravity * planet_radius : planet radius [cm] * mass : planet mass [g] * log_P_set : Pressure value to contrain the PT profile, defaults to 10 bar. * T_set : temperature at P_set to constrain the PT profile. [K] * log_P3 : (log) Pressure value for the top of the deep atmospheric layer, [bar] * P2 : in range (0,1), sets the pressure level of the middle atmospheric layer * P1 : in range (0,1), sets the pressure level of the top atmospheric layer * alpha_0 : slope of the upper atmospheric layer * alpha_1 : slope of the middle atmospheric layer Optional : * beta : power law for the slopes, default value is 0.5. Not recommended to change! Either: * log_pquench : Pressure at which CO, CH4 and H2O abundances become vertically constant * Fe/H : Metallicity * C/O : Carbon to oxygen ratio Or: * $SPECIESNAME[_$DATABASE][.R$RESOLUTION] : The log mass fraction abundance of the species Either: * [log_]Pcloud : The (log) pressure at which to place the gray cloud opacity. Or: * fsed : sedimentation parameter - can be unique to each cloud type One of: * sigma_lnorm : Width of cloud particle size distribution (log normal) * b_hans : Width of cloud particle size distribution (hansen) One of: * log_cloud_radius_* : Central particle radius (typically computed with fsed and Kzz) * log_kzz : Vertical mixing parameter One of * eq_scaling_* : Scaling factor for equilibrium cloud abundances. * log_X_cb_: cloud mass fraction abundance Optional * contribution : return the transmission contribution function * power_law_opacity_coefficient : gamma, power law slope for a rayleigh-like haze * haze_factor : multiplicative scaling factor for the strength of the rayleigh haze * power_law_opacity_350nm : strength of the rayleigh haze at 350 nm. pt_plot_mode : bool Return only the pressure-temperature profile for plotting. Evaluate mode only. adaptive_mesh_refinement : Adaptive mesh refinement. Use the high resolution pressure grid around the cloud base. Returns: wlen_model : jnp.array Wavlength array of computed model, not binned to data [um] spectrum_model : jnp.array Computed transmission spectrum planet_radius**2/Rstar**2 contr-em : ArrayLike Optional, the transmission contribution function, relative contributions for each wavelength and pressure level. Only the clear atmosphere contribution is returned. .. py:function:: legacy_madhu_seager_patchy_transmission(prt_object, parameters, pt_plot_mode=False, adaptive_mesh_refinement=False) Deprecated .. py:function:: legacy_isothermal_transmission(prt_object, parameters, pt_plot_mode=False, adaptive_mesh_refinement=False) Equilibrium Chemistry Transmission Model, Isothermal Profile This model computes a transmission spectrum based on an isothermal temperature-pressure profile. Args: prt_object : object An instance of the pRT class, with optical properties as defined in the RunDefinition. parameters : dict Dictionary of required parameters: * distance_to_system : Distance to the planet in [cm] Two of * log_g : Log of surface gravity * planet_radius : planet radius [cm] * mass : planet mass [g] * temperature : Interior temperature of the planet [K] Either: * log_pquench : Pressure at which CO, CH4 and H2O abundances become vertically constant * Fe/H : Metallicity * C/O : Carbon to oxygen ratio Or: * $SPECIESNAME[_$DATABASE][.R$RESOLUTION] : The log mass fraction abundance of the species Either: * [log_]Pcloud : The (log) pressure at which to place the gray cloud opacity. Or: * fsed : sedimentation parameter - can be unique to each cloud type One of: * sigma_lnorm : Width of cloud particle size distribution (log normal) * b_hans : Width of cloud particle size distribution (hansen) One of: * log_cloud_radius_* : Central particle radius (typically computed with fsed and Kzz) * log_kzz : Vertical mixing parameter One of * eq_scaling_* : Scaling factor for equilibrium cloud abundances. * log_X_cb_: cloud mass fraction abundance Optional * contribution : return the transmission contribution function * power_law_opacity_coefficient : gamma, power law slope for a rayleigh-like haze * haze_factor : multiplicative scaling factor for the strength of the rayleigh haze * power_law_opacity_350nm : strength of the rayleigh haze at 350 nm. pt_plot_mode : bool Return only the pressure-temperature profile for plotting. Evaluate mode only. adaptive_mesh_refinement : Adaptive mesh refinement. Use the high resolution pressure grid around the cloud base. Returns: wlen_model : jnp.array Wavlength array of computed model, not binned to data [um] spectrum_model : jnp.array Computed transmission spectrum planet_radius**2/Rstar**2 contr-em : ArrayLike Optional, the transmission contribution function, relative contributions for each wavelength and pressure level. Only the clear atmosphere contribution is returned if patchy clouds are considered. .. py:function:: legacy_power_law_profile_transmission(prt_object, parameters, pt_plot_mode=False, adaptive_mesh_refinement=False) This model computes a emission spectrum based a gradient temperature-pressure profile (Zhang 2023). Either free or equilibrium chemistry can be used, together with a range of cloud parameterizations. It is possible to use free abundances for some species and equilibrium chemistry for the remainder. Args: prt_object : object An instance of the pRT class, with optical properties as defined in the RunDefinition. parameters : dict Dictionary of required parameters: * distance_to_system : Distance to the planet in [cm] Two of * log_g : Log of surface gravity * planet_radius : planet radius [cm] * mass : planet mass [g] * alpha : power law slope for the temperture profile * T_0 : multiplicative factor for the power law slope Either: * log_pquench : Pressure at which CO, CH4 and H2O abundances become vertically constant * Fe/H : Metallicity * C/O : Carbon to oxygen ratio Or: * $SPECIESNAME[_$DATABASE][_R_$RESOLUTION] : The log mass fraction abundance of the species Optional: * fsed : sedimentation parameter - can be unique to each cloud type One of: * sigma_lnorm : Width of cloud particle size distribution (log normal) * b_hans : Width of cloud particle size distribution (hansen) One of: * log_cloud_radius_* : Central particle radius (typically computed with fsed and Kzz) * log_kzz : Vertical mixing parameter One of * eq_scaling_* : Scaling factor for equilibrium cloud abundances. * log_X_cb_: cloud mass fraction abundance Optional * contribution : return the emission contribution function * patchiness : Cloud coverage fraction, mixes two columns with different cloud properties. * remove_cloud_species : Specifies which cloud species to remove for the clear atmosphere column. * T_disk_blackbody : Temperature of a blackbody circumplanetary disk component. * disk_radius : Radius [cm] of a blackbody circumplanetary disk component. pt_plot_mode : bool Return only the pressure-temperature profile for plotting. Evaluate mode only. adaptive_mesh_refinement : Adaptive mesh refinement. Use the high resolution pressure grid around the cloud base. Returns: wlen_model : jnp.array Wavlength array of computed model, not binned to data [um] spectrum_model : jnp.array Computed transmission spectrum planet_radius**2/Rstar**2 contr-em : ArrayLike Optional, the emission contribution function, relative contributions for each wavelength and pressure level. .. py:function:: add_blackbody_cpd_model(parameters, wavelengths) Calculates the flux of a blackbody with area 4*pi*disk_radius^2 and temperature T_disk_blackbody. This is in units of W/m2/micron, and can be added to a planetary spectrum to model the contribution of a circumplanetary disk Args: parameters (dict): dictionary of atmospheric and disk parameters wavelengths (ArrayLike): Wavelength grid of atmospheric model in micron Returns: blackbody_spectrum (ArrayLike): 1D Planck emission spectrum for a circular CPD. .. py:function:: initialize_pressure(press, parameters, adaptive_mesh_refinement) .. py:function:: initialize_pressure_from_legacy_dict(press, parameters, adaptive_mesh_refinement) Legacy wrapper for :func:`initialize_pressure`. Converts a legacy parameter dictionary (values accessed via ``.value``) to a plain-value mapping before forwarding to :func:`initialize_pressure`. Args: parameters : dict Legacy parameter dictionary where each value is an object with a ``.value`` attribute. All other arguments are forwarded unchanged to :func:`initialize_pressure`. .. py:function:: _as_auxiliary_outputs(outputs) .. py:data:: _EMISSION_RUNTIME_AUXILIARY_OUTPUT_EXCLUSIONS .. py:function:: _as_filtered_auxiliary_outputs(outputs, excluded_keys=()) .. py:function:: _is_scalar_like(value) .. py:function:: _build_model_context_from_parameters(name, mode, prt_object, parameters, adaptive_mesh_refinement=False, variability_atmospheric_column_model_flux_return_mode=False, model_metadata=None) .. py:function:: _normalize_model_output(model_output, pt_plot_mode=False) .. py:function:: _model_result_to_legacy_output(model_result: petitRADTRANS.retrieval.runtime.ModelResult) .. py:function:: run_legacy_model_adapter(model_function, model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, *, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: _infer_model_function_contract(model_function, declared_contract: str) -> str Resolve the model contract using explicit registration only. Contract detection follows strict precedence: 1. Explicit ``_prt_model_contract`` attribute (set by :func:`mark_legacy_model_function`). 2. The caller-supplied *declared_contract*. 3. Fallback to :data:`MODEL_CONTRACT_LEGACY`. No signature inspection is performed — models must opt in to the runtime-native contract via the decorator or by setting the attribute directly. .. py:function:: coerce_model_generating_function(model_function, declared_contract: str = MODEL_CONTRACT_LEGACY) .. py:function:: run_runtime_model_adapter(model_function, model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params: petitRADTRANS.retrieval.runtime.PhysicalParams, *, pt_plot_mode: bool = False) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: make_unified_model(fn, legacy_fn=None) Return a single callable that dispatches to *fn* or *legacy_fn* based on the type of the first positional argument. When the first argument is a :class:`~petitRADTRANS.retrieval.runtime.ModelContext` instance, the differentiable code path is taken. Any other type routes to the legacy calling convention ``(prt_object, parameters, ...)``. If no *legacy_fn* is supplied and the first argument is not a :class:`ModelContext`, a :class:`TypeError` is raised. The dispatcher carries ``_prt_model_contract == MODEL_CONTRACT_DIFFERENTIABLE`` so that :func:`coerce_model_generating_function` and the retrieval runtime treat it as a first-class differentiable function. The constituent functions remain accessible as ``.___`` and ``.__legacy__`` attributes. .. py:function:: _initialize_runtime_pressure_grid(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params) .. py:function:: _resolve_runtime_abundance_state(model_context: petitRADTRANS.retrieval.runtime.ModelContext, physical_params, p_use, p_global, temperatures, *, temperature_builder=None, pt_plot_mode: bool = False) .. py:function:: _build_runtime_emission_model_inputs(physical_params, abundance_state, gravity, planet_radius, cloud_properties) .. py:function:: _build_runtime_transmission_model_inputs(physical_params, abundance_state, gravity, planet_radius, *, reference_pressure, opaque_cloud_top_pressure, haze_factor, power_law_opacity_coefficient, power_law_opacity_350nm, cloud_properties) .. py:function:: calculate_emission_spectrum_runtime(model_context: petitRADTRANS.retrieval.runtime.ModelContext, model_inputs: petitRADTRANS.retrieval.runtime.ModelInputs) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: calculate_emission_spectra_batched(model_context: petitRADTRANS.retrieval.runtime.ModelContext, batched_model_inputs: petitRADTRANS.retrieval.runtime.ModelInputs) -> petitRADTRANS.retrieval.runtime.ModelResult Evaluate emission spectra for a batch of model inputs via vmap. ``model_context`` (including the ``Radtrans`` object and all static flags) is treated as static across the batch. Only the per-sample arrays inside ``batched_model_inputs`` are mapped over. Each leaf of ``batched_model_inputs`` must carry a leading batch dimension. This is the preferred entry point when the caller has already constructed batched ``ModelInputs`` and wants to run the RT kernel efficiently without wrapping the full evaluation pipeline. Args: model_context: Static evaluation context shared across all samples. batched_model_inputs: A ``ModelInputs`` pytree whose leaves each have shape ``(batch_size, ...)`` — i.e. the first axis is the batch dim. Returns: A ``ModelResult`` whose ``wavelengths`` and ``spectrum`` leaves have a leading batch dimension ``(batch_size, n_wavelengths)``. .. py:function:: calculate_transmission_spectrum_runtime(model_context: petitRADTRANS.retrieval.runtime.ModelContext, model_inputs: petitRADTRANS.retrieval.runtime.ModelInputs) -> petitRADTRANS.retrieval.runtime.ModelResult .. py:function:: calculate_emission_spectrum(prt_object, parameters, temperatures, abundances, gravity, mean_molar_masses, planet_radius, sigma_lnorm, cloud_particle_mean_radii, cloud_f_sed, eddy_diffusion_coefficients, cloud_hansen_b, cloud_fraction, patchy_clouds, distribution) Calls Radtrans.calculate_flux to compute the emission spectrum of an atmosphere. This function automatically checks if patchiness is included in the retrieval, and mixes the clear and cloudy columns. Patchiness can be applied to all of the cloud species, or individual clouds can be chosen using the remove_cloud_species parameter. A circumplanetary disk model is optionally included, modelled as a blackbody with some temperature T_disk_blackbody and a radius disk_radius. Args: prt_object (Radtrans): The Radtrans object used to calculate the spectrum parameters (dict): Dictionary of atmospheric parameters. temperatures (ArrayLike): Array of temperatures for each pressure level in the atmosphere abundances (dict): Dictionary of molecular mass fraction abundances for each level in the atmosphere gravity (ArrayLike): Gravitational acceleration at each pressure level mean_molar_masses (ArrayLike): Mean molecular mass at each pressure level planet_radius (float): Planet radius in cm sigma_lnorm (float): Width of the cloud particle size distribution (log-normal) cloud_particle_mean_radii (ArrayLike): Mean particle radius cloud_f_sed (ArrayLike): Sedimentation fraction eddy_diffusion_coefficients (ArrayLike): Vertical mixing strength (Kzz) cloud_hansen_b (ArrayLike): Cloud particle distribution width, hansen distsribution cloud_fraction (float) : fraction of planet covered by clouds patchy_clouds (list(str)) : Which clouds are patchy distribution (string): Which cloud particle size distribution to use .. py:function:: calculate_transmission_spectrum(prt_object, parameters, temperatures, abundances, gravity, mean_molar_masses, planet_radius, reference_pressure, opaque_cloud_top_pressure, sigma_lnorm, cloud_particle_mean_radii, cloud_f_sed, eddy_diffusion_coefficients, haze_factor, power_law_opacity_coefficient, power_law_opacity_350nm, cloud_hansen_b, cloud_fraction, patchy_clouds, distribution) _summary_ Args: prt_object (Radtrans): The Radtrans object used to calculate the spectrum parameters (dict): Dictionary of atmospheric parameters. temperatures (ArrayLike): Array of temperatures for each pressure level in the atmosphere abundances (dict): Dictionary of molecular mass fraction abundances for each level in the atmosphere gravity (ArrayLike): Gravitational acceleration at each pressure level mean_molar_masses (ArrayLike): Mean molecular mass at each pressure level planet_radius (float):Planet radius in cm reference_pressure (float): Pressure at which the planet radius is defined opaque_cloud_top_pressure (float): Pressure where an opaque grey cloud deck is placed sigma_lnorm (float): Width of the cloud particle size distribution (log-normal) cloud_particle_mean_radii (ArrayLike): Mean particle radius cloud_f_sed (ArrayLike): Sedimentation fraction eddy_diffusion_coefficients (ArrayLike): Vertical mixing strength (Kzz) haze_factor (float): Multiplicative factor on the strength of a power law haze slope power_law_opacity_coefficient (float): Exponent for the slope of a power law haze power_law_opacity_350nm (float): Strength of the power law scattering at 350nm cloud_hansen_b (ArrayLike): Cloud particle distribution width, hansen distsribution cloud_fraction (float) : fraction of planet covered by clouds patchy_clouds (list(str)) : Which clouds are patchy distribution (string): Log normal or hansen particle size distribution Returns: _type_: _description_ .. py:data:: molliere_2020_emission .. py:data:: molliere_2020_two_column .. py:data:: guillot_emission .. py:data:: guillot_emission_add_gaussian_temperature .. py:data:: guillot_patchy_emission .. py:data:: interpolated_profile_emission .. py:data:: gradient_profile_emission .. py:data:: power_law_profile_emission .. py:data:: guillot_transmission .. py:data:: guillot_patchy_transmission .. py:data:: madhushudhan_seager_emission .. py:data:: madhushudhan_seager_transmission .. py:data:: madhu_seager_patchy_transmission .. py:data:: isothermal_transmission .. py:data:: power_law_profile_transmission .. py:data:: time_series_gradient_emission