quimb.experimental.misc.renormalization
=======================================

.. py:module:: quimb.experimental.misc.renormalization


Attributes
----------

.. autoapisummary::

   quimb.experimental.misc.renormalization.coarse_grain_


Functions
---------

.. autoapisummary::

   quimb.experimental.misc.renormalization.CTMRG
   quimb.experimental.misc.renormalization.coarse_grain_eager


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

.. py:function:: CTMRG(T, A, C, L, max_bond, strip_exponent=False)

   Contract the translationally invariant tensor network given by bulk,
   edge and corner arrays ``T``, ``A`` and ``C`` respectively with side
   length ``L``.

   :param T: The bulk tensor, dimensions corresponding to
             (towards corner, away from corner, away from corner, towards corner),
   :type T: array_like with shape (d, d, d, d)
   :param A: The edge tensor, dimensions corresponding to
             (away from corner, towards corner, towards bulk).
   :type A: array_like with shape (d, d, d)
   :param C: The corner tensor.
   :type C: array_like with shape (d, d)
   :param L: The side length of the lattice to generate.
   :type L: int
   :param max_bond: The maximum bond dimension to use.
   :type max_bond: int
   :param strip_exponent: Whether to strip an overall exponent while contracting and accrue this
                          separately, to avoid overflow.
   :type strip_exponent: bool, optional

   :returns: * **Z** (*float*) -- The contracted value of the tensor network.
             * **exponent** (*float, optional*) -- The exponent, in log10, accrued while contracting, such that the real
               contracted value is ``Z * 10**exponent``.


.. py:function:: coarse_grain_eager(self, direction, compress=True, equalize_norms=False, inplace=False, **compress_opts)

   This contracts pairs of tensors in along ``direction``, and then
   optionally compresses the doubled bonds generated along the other
   direction.


.. py:data:: coarse_grain_