cotengra.hyperoptimizers.hyper_cmaes
====================================

.. py:module:: cotengra.hyperoptimizers.hyper_cmaes

.. autoapi-nested-parse::

   Hyper parameter optimization using cmaes, as implemented by

   https://github.com/CyberAgentAILab/cmaes.



Classes
-------

.. autoapisummary::

   cotengra.hyperoptimizers.hyper_cmaes.HyperCMAESSampler
   cotengra.hyperoptimizers.hyper_cmaes.CMAESOptLib


Module Contents
---------------

.. py:class:: HyperCMAESSampler(space, sigma=1.0, lr_adapt=True, separable=False, exponential_param_power=None, **kwargs)

   .. py:attribute:: params
      :value: []



   .. py:attribute:: opt


   .. py:attribute:: _trial_counter
      :value: 0



   .. py:attribute:: _trial_store


   .. py:attribute:: _batch
      :value: []



   .. py:method:: ask()


   .. py:method:: tell(trial_number, value)


.. py:class:: CMAESOptLib

   Bases: :py:obj:`cotengra.hyperoptimizers.hyper.HyperOptLib`


   Hyper-optimization using CMA-ES with per-method optimizers and
   a Lower Confidence Bound method selector.


   .. py:method:: setup(methods, space, optimizer=None, sigma=1.0, lr_adapt=True, method_exploration=1.0, method_temperature=1.0, exponential_param_power=None, **cmaes_opts)

      Initialize the optimizer state.

      :param methods: The list of contraction methods to optimize over.
      :type methods: list[str]
      :param space: The search space for each method.
      :type space: dict[str, dict[str, dict]]
      :param optimizer: The parent ``HyperOptimizer`` instance, for accessing
                        attributes like ``max_repeats``.
      :type optimizer: HyperOptimizer, optional
      :param kwargs: Extra options specific to the optimizer library.



   .. py:method:: get_setting()

      Suggest the next setting to trial.

      :returns: **setting** -- Must contain at least ``{"method": str, "params": dict}``.
                May also include tokens for reporting.
      :rtype: dict



   .. py:method:: report_result(setting, trial, score)

      Report the result of a trial.

      :param setting: The setting dict returned by ``get_setting``.
      :type setting: dict
      :param trial: The trial result dict.
      :type trial: dict
      :param score: The scalar score for this trial.
      :type score: float