distribution_metrics
====================

.. py:module:: distribution_metrics

.. autoapi-nested-parse::

   distribution_metrics.py

   Unified module for distribution comparison, kernel metrics,
   information-theoretic losses, and entropy-regularized optimal transport.

   Callable as:
       from distribution_metrics import DistributionMetrics as DM



Classes
-------

.. autoapisummary::

   distribution_metrics.DistributionMetrics


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

.. py:class:: DistributionMetrics

   Stateless utility namespace for:
   - Kernel-based metrics (RBF, MMD)
   - Information-theoretic divergences (KL)
   - Bitstring / probability utilities
   - Sinkhorn (entropy-regularized OT) distance


   .. py:method:: rbf_kernel(x, y, gamma)
      :staticmethod:



   .. py:method:: compute_kernel_matrix(space, gammas)
      :staticmethod:



   .. py:method:: kernel_expectation(px, py, kernel_matrix)
      :staticmethod:



   .. py:method:: mmd_loss(px, py, kernel_matrix)
      :staticmethod:



   .. py:method:: kl_divergence(p, q, eps = 1e-12)
      :staticmethod:



   .. py:method:: safe_log(x, eps = 1e-12)
      :staticmethod:



   .. py:method:: normalize_probs(probs)
      :staticmethod:



   .. py:method:: int_to_bitstring(n, length)
      :staticmethod:



   .. py:method:: bitstring_to_int(bitstring)
      :staticmethod:



   .. py:method:: probs_to_bitstrings(prob_vector, threshold = 1e-06)
      :staticmethod:



   .. py:method:: compute_chi(samples, valid_bitstrings)
      :staticmethod:



   .. py:method:: sinkhorn_kernel(cost_matrix, epsilon)
      :staticmethod:



   .. py:method:: build_cost_matrix(space)
      :staticmethod:



   .. py:method:: sinkhorn_loss(p, q, space = None, epsilon = 0.05, max_iter = 200, tol = 1e-09)
      :staticmethod:



   .. py:method:: sinkhorn_report(p, q, epsilon = 0.05)
      :staticmethod: