agedi.utils
===========

.. py:module:: agedi.utils


Submodules
----------

.. toctree::
   :maxdepth: 1

   /autoapi/agedi/utils/offsets/index
   /autoapi/agedi/utils/truncated_normal/index


Attributes
----------

.. autoapisummary::

   agedi.utils.OFFSET_LIST


Classes
-------

.. autoapisummary::

   agedi.utils.TruncatedNormal


Package Contents
----------------

.. py:class:: TruncatedNormal(loc: Union[numbers.Number, torch.Tensor], scale: Union[numbers.Number, torch.Tensor], a: Union[numbers.Number, torch.Tensor], b: Union[numbers.Number, torch.Tensor], validate_args: Optional[bool] = None)

   Bases: :py:obj:`TruncatedStandardNormal`


   Truncated Normal distribution
   https://people.sc.fsu.edu/~jburkardt/presentations/truncated_normal.pdf


   .. py:attribute:: has_rsample
      :value: True



   .. py:attribute:: _log_scale


   .. py:attribute:: _mean


   .. py:attribute:: _variance


   .. py:method:: _to_std_rv(value: torch.Tensor) -> torch.Tensor

      Standardise *value* to the standard (zero-mean, unit-variance) domain.



   .. py:method:: _from_std_rv(value: torch.Tensor) -> torch.Tensor

      Map *value* from the standard domain back to the original (loc/scale) domain.



   .. py:method:: cdf(value: torch.Tensor) -> torch.Tensor

      Cumulative distribution function evaluated at *value*.



   .. py:method:: icdf(value: torch.Tensor) -> torch.Tensor

      Inverse CDF (quantile function) evaluated at *value*.



   .. py:method:: log_prob(value: torch.Tensor) -> torch.Tensor

      Log probability density evaluated at *value*.



.. py:data:: OFFSET_LIST