emmi.pipeline.batch_processors.position_normalization
=====================================================

.. py:module:: emmi.pipeline.batch_processors.position_normalization


Classes
-------

.. autoapisummary::

   emmi.pipeline.batch_processors.position_normalization.PositionNormalizationBatchProcessor


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

.. py:class:: PositionNormalizationBatchProcessor(items, raw_pos_min, raw_pos_max, scale = 1000)

   Bases: :py:obj:`ksuit.data.pipeline.BatchProcessor`


   Post-processes data on a batch-level to normalize positions.

   Initializes the PositionNormalizationPostCollator

   :param items: The position items to normalize.
   :param raw_pos_min: The minimum position in the source domain.
   :param raw_pos_max: The maximum position in the source domain.
   :param scale: The maximum value of the position. Defaults to 1000.


   .. py:attribute:: items


   .. py:attribute:: scale
      :value: 1000



   .. py:attribute:: raw_pos_min_tensor


   .. py:attribute:: raw_pos_max_tensor


   .. py:attribute:: raw_size


   .. py:method:: denormalize(key, value)

      Inverts the normalization from the __call__ method of a single item in the batch.

      :param key: The name of the item.
      :param value: The value of the item.

      :returns: The same name and the denormalized value.
      :rtype: (key, value)