`leaspy.algo.fit.abstract_mcmc`.AbstractFitMCMC

class AbstractFitMCMC(settings)

Bases: AlgoWithAnnealingMixin, AlgoWithSamplersMixin, AbstractFitAlgo

Abstract class containing common method for all fit algorithm classes based on Monte-Carlo Markov Chains (MCMC).

Parameters:

settingsAlgorithmSettings: MCMC fit algorithm settings

See also

leaspy.algo.utils.samplers

Attributes:

samplersdict[ str, AbstractSampler ]: Dictionary of samplers per each variable
random_order_variablesbool (default True): This attribute controls whether we randomize the order of variables at each iteration. Article https://proceedings.neurips.cc/paper/2016/hash/e4da3b7fbbce2345d7772b0674a318d5-Abstract.html gives a rationale on why we should activate this flag.
temperaturefloat
temperature_invfloat: Temperature and its inverse (modified during algorithm when using annealing)

Methods

`iteration`(dataset, model, realizations)	MCMC-SAEM iteration.
`load_parameters`(parameters)	Update the algorithm's parameters by the ones in the given dictionary.
`run`(model, *args[, return_loss])	Main method, run the algorithm.
`run_impl`(model, dataset)	Main method, run the algorithm.
`set_output_manager`(output_settings)	Set a `FitOutputManager` object for the run of the algorithm

iteration(dataset: Dataset, model: AbstractModel, realizations: CollectionRealization) → None

MCMC-SAEM iteration.

Sample : MC sample successively of the population and individual variables
Maximization step : update model parameters from current population/individual variables values.

Parameters:

datasetDataset
modelAbstractModel
realizationsCollectionRealization

load_parameters(parameters: dict)

Update the algorithm’s parameters by the ones in the given dictionary. The keys in the io which does not belong to the algorithm’s parameters keys are ignored.

Parameters:

parametersdict: Contains the pairs (key, value) of the wanted parameters

Examples

>>> settings = leaspy.io.settings.algorithm_settings.AlgorithmSettings("mcmc_saem")
>>> my_algo = leaspy.algo.fit.tensor_mcmcsaem.TensorMCMCSAEM(settings)
>>> my_algo.algo_parameters
{'n_iter': 10000,
 'n_burn_in_iter': 9000,
 'eps': 0.001,
 'L': 10,
 'sampler_ind': 'Gibbs',
 'sampler_pop': 'Gibbs',
 'annealing': {'do_annealing': False,
  'initial_temperature': 10,
  'n_plateau': 10,
  'n_iter': 200}}
>>> parameters = {'n_iter': 5000, 'n_burn_in_iter': 4000}
>>> my_algo.load_parameters(parameters)
>>> my_algo.algo_parameters
{'n_iter': 5000,
 'n_burn_in_iter': 4000,
 'eps': 0.001,
 'L': 10,
 'sampler_ind': 'Gibbs',
 'sampler_pop': 'Gibbs',
 'annealing': {'do_annealing': False,
  'initial_temperature': 10,
  'n_plateau': 10,
  'n_iter': 200}}

run(model: AbstractModel, *args, return_loss: bool = False, **extra_kwargs) → Any

Main method, run the algorithm.

TODO fix proper abstract class method: input depends on algorithm… (esp. simulate != from others…)

Parameters:

modelAbstractModel: The used model.
datasetDataset: Contains all the subjects’ observations with corresponding timepoints, in torch format to speed up computations.
return_lossbool (default False), keyword only: Should the algorithm return main output and optional loss output as a 2-tuple?

Returns:

Depends on algorithm class: TODO change?

See also

AbstractFitAlgo
AbstractPersonalizeAlgo
SimulationAlgorithm

run_impl(model: AbstractModel, dataset: Dataset)

Main method, run the algorithm.

Basically, it initializes the CollectionRealization object, updates it using the iteration method then returns it.

TODO fix proper abstract class

Parameters:

modelAbstractModel: The used model.
datasetDataset: Contains the subjects’ observations in torch format to speed up computation.

Returns:

2-tuple:

realizationsCollectionRealization
The optimized parameters.
None : placeholder for noise-std

set_output_manager(output_settings: OutputsSettings) → None

Set a FitOutputManager object for the run of the algorithm

Parameters:

output_settingsOutputsSettings: Contains the logs settings for the computation run (console print periodicity, plot periodicity …)

Examples

>>> from leaspy import AlgorithmSettings
>>> from leaspy.io.settings.outputs_settings import OutputsSettings
>>> from leaspy.algo.fit.tensor_mcmcsaem import TensorMCMCSAEM
>>> algo_settings = AlgorithmSettings("mcmc_saem")
>>> my_algo = TensorMCMCSAEM(algo_settings)
>>> settings = {'path': 'brouillons',
                'console_print_periodicity': 50,
                'plot_periodicity': 100,
                'save_periodicity': 50
                }
>>> my_algo.set_output_manager(OutputsSettings(settings))

leaspy.algo.fit.abstract_mcmc.AbstractFitMCMC

`leaspy.algo.fit.abstract_mcmc`.AbstractFitMCMC