clr_samples_to_vmr#

exo_skryer.build_chem.clr_samples_to_vmr(samples_dict: dict[str, 'np.ndarray'], species: tuple[str, ...] | list[str]) → dict[str, 'np.ndarray'][source]#

Convert CLR posterior samples to physical VMR (and log10 VMR) columns.

Applies the same softmax inverse as constant_vmr_clr but operates on NumPy arrays of posterior samples rather than JAX scalars.

Parameters:
samples_dictdict[str, np.ndarray]

Posterior samples keyed by parameter name. Must contain clr_<species> for every species in species.

speciestuple or list of str

Ordered trace species names (e.g. ('H2O', 'CO', 'CO2')).

Returns:
deriveddict[str, np.ndarray]

New columns ready to merge into the samples dictionary:

  • log_10_f_<species> : log10(VMR) for each trace species

  • f_<species> : linear VMR for each trace species

  • f_H2_He : combined H2+He filler fraction