This Python package is an implementation of the Multilevel Generalized Likelihood Uncertainty Estimation (MLGLUE) algorithm, including some utility functions. See the MLGLUE paper here (preprint).
pip install MLGLUE
MLGLUE uses Ray for parallelization. Installation using pip install MLGLUE also installs Ray as dependency along numpy and matplotlib. Using a custom installation, please prepare an environment for Ray first as described here for version 2.2.0. The use of other versions of Ray has not been tested.
A documentation webpage is available here or alternatively under mlglue.readthedocs.io.
MLGLUE requires the computational model to be given in the form of a function with a specific set of arguments and returns:
def my_model(parameters, level, n_levels, obs_x, obs_y, likelihood, run_id)
"""
Parameters
----------
# :param parameters: 1D list-like of model parameters
# :param level: int representing the 0-based level index
# :param n_levels: int representing the total number of levels
# :param obs_x: 1D list-like of virtual observation ordinates
# (not actually used in computations)
# :param obs_y: 1D list-like of observations
# :param likelihood: an instance of an MLGLUE.Likelihoods
# likelihood instance
# :param run_id: an int or str identifier for the model run
Returns
-------
# :return computed_likelihood: the computed likelihood as float
# :return simulated_observation_equivalents: the simulated
# equivalents of the observations
"""
# your model code
# using the function parameter_to_observable_map as a placeholder
# for more complex models
simulated_observation_equivalents = parameter_to_observable_map(
parameters
)
# compute likelihood
computed_likelihood = likelihood.likelihood(
obs=obs_y,
sim=simulated_observation_equivalents
)
return computed_likelihood, simulated_observation_equivalentsSee the examples directory for a working implementation.