Tensor

class adcc.Tensor(sym_or_mo, space=None, permutations=None, spin_block_maps=None, spin_blocks_forbidden=None)

Bases: libadcc.Tensor

Construct an uninitialised Tensor from an MoSpaces or a Symmetry object.

More information about the last four, symmetry-related parameters see the documentation of the Symmetry object.

Parameters

  • sym_or_mo – Symmetry or MoSpaces object

  • spaces (str, optional) – Space of the tensor, can be None if the first argument is a Symmetry object.

  • permutations (list, optional) – List of permutational symmetries of the Tensor.

  • spin_block_maps (list, optional) – List of mappings between spin blocks

  • spin_blocks_forbidden (list, optional) – List of forbidden (i.e. forced-to-zero) spin blocks.

Notes

An MoSpaces object is contained in many datastructures of adcc, including the AdcMatrix, the LazyMp, the ReferenceState and any solver or ADC results state.

Examples

Construct a symmetric tensor in the “o1o1” (occupied-occupied) spaces:

>>> Tensor(mospaces, "o1o1", permutations=["ij", "ji"])

Construct an anti-symmetric tensor in the “v1v1” spaces:

>>> Tensor(mospaces, "v1v1", permutations=["ab", "-ba"])

Construct a tensor in “o1v1”, which maps the alpha-alpha block anti-symmetrically to the beta-beta block and which has the other spin blocks set to zero:

>>> Tensor(mospaces, "o1v1", spin_block_maps=[("aa", "bb", -1)],
...        spin_blocks_forbidden=["ab", "ba"])

Methods Summary

select_below_absmax(tolerance)

Select the absolute maximal values in the tensor, which are below the given tolerance.

Methods Documentation

select_below_absmax(tolerance)

Select the absolute maximal values in the tensor, which are below the given tolerance.