leaspy.samplers.base
====================

.. py:module:: leaspy.samplers.base


Classes
-------

.. autoapisummary::

   leaspy.samplers.base.AbstractSampler
   leaspy.samplers.base.AbstractIndividualSampler
   leaspy.samplers.base.AbstractPopulationSampler


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

.. py:class:: AbstractSampler(name, shape, *, acceptation_history_length = 25)

   Bases: :py:obj:`abc.ABC`


   
   Abstract sampler class.


   :Parameters:

       **name** : :obj:`str`
           The name of the random variable to sample.

       **shape** : :obj:`tuple` of :obj:`int`
           The shape of the random variable to sample.

       **acceptation_history_length** : :obj:`int` > 0 (default 25)
           Deepness (= number of iterations) of the history kept for computing the mean acceptation rate.
           (It is the same for population or individual variables.)

   :Attributes:

       **name** : :obj:`str`
           Name of variable

       **shape** : :obj:`tuple` of :obj:`int`
           Shape of variable

       **acceptation_history_length** : :obj:`int`
           Deepness (= number of iterations) of the history kept for computing the mean acceptation rate.
           (Same for population or individual variables by default.)

       **acceptation_history** : :class:`torch.Tensor`
           History of binary acceptations to compute mean acceptation rate for the sampler in MCMC-SAEM algorithm.
           It keeps the history of the last `acceptation_history_length` steps.






   :Raises:

       :exc:`.LeaspyModelInputError`
           ..







   ..
       !! processed by numpydoc !!

   .. py:attribute:: name


   .. py:attribute:: shape


   .. py:attribute:: acceptation_history_length
      :value: 25



   .. py:attribute:: acceptation_history


   .. py:property:: ndim
      :type: int


      
      Number of dimensions.
















      ..
          !! processed by numpydoc !!


   .. py:property:: shape_acceptation
      :type: Tuple[int, Ellipsis]

      :abstractmethod:


      
      Return the shape of acceptation tensor for a single iteration.





      :Returns:

          :obj:`tuple` of :obj:`int`
              The shape of the acceptation history.











      ..
          !! processed by numpydoc !!


   .. py:method:: sample(state, *, temperature_inv)
      :abstractmethod:


      
      Apply a sampling step

      <!> It will modify in-place the internal state, caching all intermediate values needed to efficient.

      :Parameters:

          **state** : :class:`.State`
              Object containing values for all model variables, including latent variables

          **temperature_inv** : :obj:`float` > 0
              Inverse of the temperature used in tempered MCMC-SAEM














      ..
          !! processed by numpydoc !!


.. py:class:: AbstractIndividualSampler(name, shape, *, n_patients, acceptation_history_length = 25)

   Bases: :py:obj:`AbstractSampler`


   
   Abstract class for samplers of individual random variables.


   :Parameters:

       **name** : :obj:`str`
           The name of the random variable to sample.

       **shape** : :obj:`tuple` of :obj:`int`
           The shape of the random variable to sample.

       **n_patients** : :obj:`int`
           Number of patients.

       **acceptation_history_length** : :obj:`int` > 0 (default 25)
           Deepness (= number of iterations) of the history kept for computing the mean acceptation rate.
           (It is the same for population or individual variables.)

   :Attributes:

       **name** : :obj:`str`
           Name of variable

       **shape** : :obj:`tuple` of :obj:`int`
           Shape of variable

       **n_patients** : :obj:`int`
           Number of patients.

       **acceptation_history_length** : :obj:`int`
           Deepness (= number of iterations) of the history kept for computing the mean acceptation rate.
           (It is the same for population or individual variables.)

       **acceptation_history** : :class:`torch.Tensor`
           History of binary acceptations to compute mean acceptation rate for the sampler in MCMC-SAEM algorithm.
           It keeps the history of the last `acceptation_history_length` steps.













   ..
       !! processed by numpydoc !!

   .. py:attribute:: n_patients


.. py:class:: AbstractPopulationSampler(name, shape, *, acceptation_history_length = 25, mask = None)

   Bases: :py:obj:`AbstractSampler`


   
   Abstract class for samplers of population random variables.


   :Parameters:

       **name** : :obj:`str`
           The name of the random variable to sample.

       **shape** : :obj:`tuple` of :obj:`int`
           The shape of the random variable to sample.

       **acceptation_history_length** : :obj:`int` > 0 (default 25)
           Deepness (= number of iterations) of the history kept for computing the mean acceptation rate.
           (It is the same for population or individual variables.)

       **mask** : :class:`torch.Tensor`, optional
           A binary (0/1) tensor indicating which elements to sample.
           Elements with value 1 (True) are included in the sampling; elements with 0 (False) are excluded.

   :Attributes:

       **name** : :obj:`str`
           Name of variable

       **shape** : :obj:`tuple` of :obj:`int`
           Shape of variable

       **acceptation_history_length** : :obj:`int`
           Deepness (= number of iterations) of the history kept for computing the mean acceptation rate.
           (It is the same for population or individual variables.)

       **acceptation_history** : :class:`torch.Tensor`
           History of binary acceptations to compute mean acceptation rate for the sampler in MCMC-SAEM algorithm.
           It keeps the history of the last `acceptation_history_length` steps.

       **mask** : :class:`torch.Tensor` of :obj:`bool`, optional
           A binary (0/1) tensor indicating which elements to sample.
           Elements with value 1 (True) are included in the sampling; elements with 0 (False) are excluded.













   ..
       !! processed by numpydoc !!

   .. py:attribute:: mask
      :value: None