gaussian.experiment
===================

.. py:module:: gaussian.experiment

.. autoapi-nested-parse::

   What is trained in the Gaussian study, and early validation of the training settings.



Classes
-------

.. autoapisummary::

   gaussian.experiment.GaussianExperiment


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

.. py:class:: GaussianExperiment

   Bases: :py:obj:`nnpd.Experiment`


   What is trained: problem, prior, training data, model and training procedure.

   Selected by "experiment" in the training settings. The runner owns execution and files.


   .. py:attribute:: name
      :value: 'gaussian-prior-dependence'



   .. py:attribute:: version
      :value: '1'



   .. py:method:: make_problem(settings)


   .. py:method:: make_prior(settings, member, problem)


   .. py:method:: members(settings)

      A fixed comparison cohort within each sweep configuration.



   .. py:method:: sample_training(context, writer)


   .. py:method:: build_model(context)


   .. py:method:: train(context, model)

      Modify model in place; return JSON-serializable training history.



   .. py:method:: signature(stage, context)

      Override for cross-run sharing; conservative default uses all training settings.



   .. py:method:: sources(stage)

      Include modules implementing transitive scientific dependencies.



   .. py:method:: validate(settings)

      Fail before any sampling or training, including for invalid alternatives.



   .. py:method:: estimate(settings, member)

      Training workload shown by the plan.