1.3. tnpy.matrix_product_state.Environment

class tnpy.matrix_product_state.Environment(mpo, mps)

Bases: object

The effective environment for MatrixProductState-based algorithms.

Parameters

• mpo (MatrixProductOperator) – Input MatrixProductOperator.

• mps (MatrixProductState) – Input MatrixProductState.

__init__(mpo, mps)

The effective environment for MatrixProductState-based algorithms.

Parameters

• mpo (tnpy.operators.MatrixProductOperator) – Input MatrixProductOperator.

• mps (tnpy.matrix_product_state.MatrixProductState) – Input MatrixProductState.

Methods

__init__(mpo, mps)	The effective environment for MatrixProductState-based algorithms.
close()	Delete the underlying data of left and right environments.
one_site_full_matrix(site)	Construct the effective matrix for variational solver, with one site remained un-contracted.
one_site_matvec(site)	Construct the effective linear operator (matrix-vector product) for variational solver, with one site remained un-contracted.
split_tensor(site, direction)
update(site, direction)	Alias to update_left() or update_right() depending on the given direction.
update_left(site)	Update the left environment at site.
update_mps(site, data)
update_right(site)	Update the right environment at site.
variance()

Attributes

left	Get a dictionary where left environments are stored on the respective site.
mpo
mps
n_sites
right	Get a dictionary where right environments are stored on the respective site.

__init__(mpo, mps)

The effective environment for MatrixProductState-based algorithms.

Parameters

• mpo (tnpy.operators.MatrixProductOperator) – Input MatrixProductOperator.

• mps (tnpy.matrix_product_state.MatrixProductState) – Input MatrixProductState.

close()

Delete the underlying data of left and right environments.

Returns:

property mpo: tnpy.operators.MatrixProductOperator

property mps: tnpy.matrix_product_state.MatrixProductState

property n_sites: int

property left: Dict[int, quimb.tensor.tensor_core.Tensor]

Get a dictionary where left environments are stored on the respective site.

Returns:

property right: Dict[int, quimb.tensor.tensor_core.Tensor]

Get a dictionary where right environments are stored on the respective site.

Returns:

update_left(site)

Update the left environment at site.

Parameters

site (int) –

Returns:

update_right(site)

Update the right environment at site.

Parameters

site (int) –

Returns:

update(site, direction)

Alias to update_left() or update_right() depending on the given direction.

Parameters

• site (int) –

• direction (tnpy.matrix_product_state.Direction) –

Returns:

update_mps(site, data)

Parameters

• site (int) –

• data (numpy.ndarray) –

split_tensor(site, direction)

Parameters

• site (int) –

• direction (tnpy.matrix_product_state.Direction) –

variance()

Return type

float

one_site_full_matrix(site)

Construct the effective matrix for variational solver, with one site remained un-contracted.

Parameters

site (int) –

Return type

numpy.ndarray

Returns:

one_site_matvec(site)

Construct the effective linear operator (matrix-vector product) for variational solver, with one site remained un-contracted.

Parameters

site (int) –

Return type

scipy.sparse.linalg._interface.LinearOperator

Returns: