pmrf.fitting.BayesianContext

class pmrf.fitting.BayesianContext(model, measured, frequency, features, measured_features, output_path=None, output_root=None, sparam_kind=None, logger=None, likelihood_kind=None, likelihood_params=None, feature_sigmas=None)

Bases: FitContext

Context object for Bayesian fitting, containing likelihood configurations.

Parameters:

• model (Model)

• measured (Network | NetworkCollection)

• frequency (Frequency)

• features (list[tuple[str, str, tuple[int, int]]])

• measured_features (ndarray)

• output_path (str | None)

• output_root (str | None)

• sparam_kind (str | None)

• logger (Logger | None)

• likelihood_kind (str)

• likelihood_params (dict[str, Parameter])

• feature_sigmas (list[str] | None)

likelihood_kind

The type of likelihood function (e.g., ‘gaussian’, ‘multivariate_gaussian’).

Type:

str or None

likelihood_params

Parameters governing the likelihood function (e.g., noise standard deviation).

Type:

dict[str, Parameter] or None

feature_sigmas

Mapping of feature names to likelihood parameter names for multivariate cases.

Type:

list[str] or None

__init__(model, measured, frequency, features, measured_features, output_path=None, output_root=None, sparam_kind=None, logger=None, likelihood_kind=None, likelihood_params=None, feature_sigmas=None)

Parameters:

• model (Model)

• measured (Network | NetworkCollection)

• frequency (Frequency)

• features (list[tuple[str, str, tuple[int, int]]])

• measured_features (ndarray)

• output_path (str | None)

• output_root (str | None)

• sparam_kind (str | None)

• logger (Logger | None)

• likelihood_kind (str | None)

• likelihood_params (dict[str, Parameter] | None)

• feature_sigmas (list[str] | None)

Return type:

None

Methods

__delattr__(name, /)	Implement delattr(self, name).
__dir__()	Default dir() implementation.
__eq__(other)	Return self==value.
__format__(format_spec, /)	Default object formatter.
__ge__(value, /)	Return self>=value.
__getattribute__(name, /)	Return getattr(self, name).
__gt__(value, /)	Return self>value.
__init__(model, measured, frequency, ...[, ...])
__init_subclass__	This method is called when a class is subclassed.
__le__(value, /)	Return self<=value.
__lt__(value, /)	Return self<value.
__ne__(value, /)	Return self!=value.
__new__(**kwargs)
__reduce__()	Helper for pickle.
__reduce_ex__(protocol, /)	Helper for pickle.
__repr__()	Return repr(self).
__setattr__(name, value, /)	Implement setattr(self, name, value).
__sizeof__()	Size of object in memory, in bytes.
__str__()	Return str(self).
__subclasshook__	Abstract classes can override this to customize issubclass().
flat_param_names()	Get names of all parameters (model and likelihood).
likelihood_param_names()	Get names of the likelihood parameters.
make_feature_function([as_numpy])	Create a JIT-compiled function to extract features from model parameters.
make_gaussian_log_likelihood_fn()	Create a simple Gaussian log-likelihood function (single sigma).
make_log_likelihood_fn([as_numpy])	Create the log-likelihood function based on the context configuration.
make_log_prior_fn([as_numpy])	Create the log-prior probability density function.
make_multivariate_gaussian_log_likelihood_fn()	Create a multivariate Gaussian log-likelihood function (multiple sigmas).
make_prior_transform_fn([as_numpy])	Create the prior transform function (unit hypercube to parameter space).
model_param_names()	Get the names of the flat parameters of the model.
settings([solver_kwargs, fitter_kwargs])	Create a FitSettings object from the current context.

Attributes

__annotations__
__dataclass_fields__
__dataclass_params__
__dict__
__doc__
__hash__
__match_args__
__module__
__weakref__	list of weak references to the object (if defined)
feature_sigmas
likelihood_kind
likelihood_params
logger
num_likelihood_params	Number of parameters in the likelihood function.
num_model_params	Number of flat parameters in the model.
num_params	Total number of parameters (model + likelihood).
output_path
output_root
sparam_kind
model
measured
frequency
features
measured_features

likelihood_kind: str = None

likelihood_params: dict[str, Parameter] = None

feature_sigmas: list[str] | None = None

property num_params: int

Total number of parameters (model + likelihood).

Type:

int

property num_model_params: int

Number of flat parameters in the model.

Type:

int

property num_likelihood_params: int

Number of parameters in the likelihood function.

Type:

int

likelihood_param_names()

Get names of the likelihood parameters.

Returns:

The names of the likelihood parameters.

Return type:

list of str

flat_param_names()

Get names of all parameters (model and likelihood).

Returns:

Combined list of parameter names.

Return type:

list of str

make_prior_transform_fn(as_numpy=False)

Create the prior transform function (unit hypercube to parameter space).

Parameters:

as_numpy (bool, optional, default=False) – If True, returns a function compatible with NumPy arrays; otherwise JAX arrays.

Returns:

Function transforming a unit hypercube vector u to parameter vector theta.

Return type:

callable

make_log_prior_fn(as_numpy=False)

Create the log-prior probability density function.

Parameters:

as_numpy (bool, optional, default=False) – If True, returns a function compatible with NumPy arrays.

Returns:

Function returning the log-probability of the given parameters.

Return type:

callable

make_log_likelihood_fn(as_numpy=False)

Create the log-likelihood function based on the context configuration.

Parameters:

as_numpy (bool, optional, default=False) – If True, returns a function compatible with NumPy arrays.

Returns:

Function returning the log-likelihood of the parameters given the data.

Return type:

callable

Raises:

Exception – If the likelihood_kind is unsupported.

make_gaussian_log_likelihood_fn()

Create a simple Gaussian log-likelihood function (single sigma).

Returns:

JIT-compiled log-likelihood function.

Return type:

callable

make_multivariate_gaussian_log_likelihood_fn()

Create a multivariate Gaussian log-likelihood function (multiple sigmas).

Returns:

JIT-compiled log-likelihood function.

Return type:

callable

__init__(model, measured, frequency, features, measured_features, output_path=None, output_root=None, sparam_kind=None, logger=None, likelihood_kind=None, likelihood_params=None, feature_sigmas=None)

Parameters:

• model (Model)

• measured (Network | NetworkCollection)

• frequency (Frequency)

• features (list[tuple[str, str, tuple[int, int]]])

• measured_features (ndarray)

• output_path (str | None)

• output_root (str | None)

• sparam_kind (str | None)

• logger (Logger | None)

• likelihood_kind (str | None)

• likelihood_params (dict[str, Parameter] | None)

• feature_sigmas (list[str] | None)

Return type:

None