MCResult Namespace Reference

Classes
class	MCConfig
class	MCObservableCovariance
class	MCRealization
class	MCResult

Functions
	_require (value, typ, str name)
	_cpp_binned_observable_id (BinnedObservableId obs)
	_cpp_param_id (ParamId pid)
ObsSample	_obs_sample_from_cpp (row)
	_obs_sample_to_cpp (Mapping[BinnedObservableId, float] row)
NuisanceSample	_nuisance_sample_from_cpp (row)
	_nuisance_sample_to_cpp (Mapping[ParamId, float] row)
List[BinnedObservableId]	covariance_ids_from_first_sample (Sequence[Mapping[BinnedObservableId, float]] samples)
MCObservableCovariance	covariance_from_obs_samples (Sequence[Mapping[BinnedObservableId, float]] samples, Sequence[BinnedObservableId] ids, float ridge_rel=1e-8, float ridge_abs=1e-12)

Detailed Description

Monte-Carlo result wrappers for the statistic module.

The C++ statistic layer can sample nuisance parameters, propagate them through
the observable model, summarize the resulting observable samples and estimate an
observable covariance matrix. This module exposes Python dataclasses mirroring
those C++ result structures.

Function Documentation

_cpp_binned_observable_id()

MCResult._cpp_binned_observable_id ( BinnedObservableId  obs )

protected

Convert a Python binned observable id to C++.

Definition at line 41 of file MCResult.py.

_cpp_param_id()

MCResult._cpp_param_id ( ParamId  pid )

protected

Convert a Python parameter id to C++.

Definition at line 46 of file MCResult.py.

_nuisance_sample_from_cpp()

NuisanceSample MCResult._nuisance_sample_from_cpp (   row )

protected

Convert one C++ nuisance sample row to a Python mapping.

Definition at line 61 of file MCResult.py.

_nuisance_sample_to_cpp()

MCResult._nuisance_sample_to_cpp ( Mapping[ParamId, float]  row )

protected

Convert one Python nuisance sample row to a C++ mapping.

Definition at line 66 of file MCResult.py.

_obs_sample_from_cpp()

ObsSample MCResult._obs_sample_from_cpp (   row )

protected

Convert one C++ observable sample row to a Python mapping.

Definition at line 51 of file MCResult.py.

_obs_sample_to_cpp()

MCResult._obs_sample_to_cpp ( Mapping[BinnedObservableId, float]  row )

protected

Convert one Python observable sample row to a C++ mapping.

Definition at line 56 of file MCResult.py.

_require()

MCResult._require (   value,   typ, str  name  )

protected

Validate a wrapper argument type and return the original value.

Definition at line 34 of file MCResult.py.

covariance_from_obs_samples()

MCObservableCovariance MCResult.covariance_from_obs_samples	(	Sequence[Mapping[BinnedObservableId, float]]	samples,
		Sequence[BinnedObservableId]	ids,
		float	ridge_rel = 1e-8,
		float	ridge_abs = 1e-12
	)

Estimate an observable covariance matrix from Monte-Carlo samples.

Args:
    samples: Observable samples produced by Monte-Carlo propagation.
    ids: Observable ordering to use for the covariance matrix.
    ridge_rel: Relative diagonal ridge added before matrix inversion.
    ridge_abs: Absolute minimum diagonal ridge.

Returns:
    A ``MCObservableCovariance`` containing the mean, covariance and inverse
    covariance.

Examples:
    >>> ids = covariance_ids_from_first_sample(samples)
    >>> cov = covariance_from_obs_samples(samples, ids)
    >>> len(cov.ids) == len(cov.mean)
    True

Definition at line 240 of file MCResult.py.

covariance_ids_from_first_sample()

List[BinnedObservableId] MCResult.covariance_ids_from_first_sample

(

Sequence[Mapping[BinnedObservableId, float]]

samples

)

Return the observable ordering inferred from the first sample.

Args:
    samples: Sequence of observable sample mappings.

Returns:
    Observable identifiers in the backend order inferred from the first row.

Raises:
    Exception: Propagates backend errors, for example when ``samples`` is
        empty.

Definition at line 218 of file MCResult.py.