tensorcircuit.zx.converter

Converter from TensorCircuit to PyZX. Includes graph preparation and reduction utilities.

class tensorcircuit.zx.converter.GraphRepresentation(graph: 'GraphS' = <factory>, rec: 'List[int]' = <factory>, silent_rec: 'List[int]' = <factory>, detectors: 'List[int]' = <factory>, observables_dict: 'Dict[int, int]'=<factory>, first_vertex: 'Dict[int, int]'=<factory>, last_vertex: 'Dict[int, int]'=<factory>, channel_probs: 'List[Any]' = <factory>, correlated_error_probs: 'List[float]' = <factory>, num_error_bits: 'int' = 0, num_correlated_error_bits: 'int' = 0)

Bases: object

__init__(graph: GraphS = <factory>, rec: List[int] = <factory>, silent_rec: List[int] = <factory>, detectors: List[int] = <factory>, observables_dict: Dict[int, int]=<factory>, first_vertex: Dict[int, int]=<factory>, last_vertex: Dict[int, int]=<factory>, channel_probs: List[Any] = <factory>, correlated_error_probs: List[float] = <factory>, num_error_bits: int = 0, num_correlated_error_bits: int = 0) → None

add_edge(e: Any, t: Any = EdgeType.SIMPLE) → None

add_edge_table(etab: dict[tuple[Any, Any], list[int]]) → None

add_edges(edges: list[tuple[Any, Any]], t: Any = EdgeType.SIMPLE) → None

add_to_phase(v: Any, p: Any, params: Any = None) → None

add_vertex(t: Any = VertexType.Z, qubit: int = -1, row: float = -1, phase: Any = 0) → int

Add a vertex to the graph.

Parameters:

• t (Any, optional) – Vertex type, defaults to VertexType.Z.

• qubit (int, optional) – Qubit index, defaults to -1.

• row (float, optional) – Row index for layout, defaults to -1.

• phase (Any, optional) – Vertex phase, defaults to 0.

Returns:

The index of the new vertex.

Return type:

int

channel_probs: List[Any]

copy() → GraphRepresentation

correlated_error_probs: List[float]

detectors: List[int]

edge(v1: Any, v2: Any) → Any

edge_set() → Any

edge_st(e: Any) → Any

Get the endpoints of an edge.

Parameters:

e (Any) – Edge.

Returns:

Tuple of vertex indices (v1, v2).

Return type:

Any

edge_type(e: Any) → Any

edges(*args: Any) → Any

first_vertex: Dict[int, int]

fuse_phases(v1: Any, v2: Any) → None

get_auto_simplify() → bool

get_params(v: Any) → Any

graph: GraphS

incident_edges(v: Any) → Any

Get edges incident to a vertex.

Parameters:

v (Any) – Vertex index.

Returns:

Iterable of edges.

Return type:

Any

inputs() → Any

is_ground(v: Any) → bool

is_multigraph() → bool

last_vertex: Dict[int, int]

property merge_vdata: Any

neighbors(v: Any) → Any

num_correlated_error_bits: int = 0

num_edges() → int

num_error_bits: int = 0

num_vertices() → int

property observables: list[int]

List of observable vertices.

Returns:

List of vertex indices.

Return type:

list[int]

observables_dict: Dict[int, int]

outputs() → Any

phase(v: Any) → Any

phases() → Any

qubit(v: Any) → Any

qubits() → Any

rec: List[int]

remove_edge(e: Any) → None

remove_edges(edges: list[tuple[Any, Any]]) → None

remove_isolated_vertices() → None

remove_vertex(v: Any) → None

remove_vertices(vertices: list[Any]) → None

row(v: Any) → Any

rows() → Any

property scalar: Any

set_auto_simplify(v: bool) → None

set_edge_type(e: Any, t: Any) → None

set_ground(v: Any, g: bool = True) → None

set_inputs(v: Any) → None

set_outputs(v: Any) → None

set_phase(v: Any, p: Any) → None

set_qubit(v: Any, q: Any) → None

set_row(v: Any, r: Any) → None

set_type(v: Any, t: Any) → None

set_vdata(v: Any, key: str, val: Any) → None

silent_rec: List[int]

to_tensor() → Any

property track_phases: bool

type(v: Any) → Any

types() → Any

update_phase_index(v1: Any, v2: Any) → None

vdata(v: Any, key: str) → Any

vdata_keys(v: Any) → Any

vertex_degree(v: Any) → int

vertex_set() → Any

vertices() → Any

class tensorcircuit.zx.converter.SamplingGraph(graph: 'Any', error_transform: 'Any', channel_probs: 'list[Any]', num_outputs: 'int', num_detectors: 'int', num_error_bits: 'int', observables: 'list[int]' = <factory>)

Bases: object

__init__(graph: ~typing.Any, error_transform: ~typing.Any, channel_probs: list[~typing.Any], num_outputs: int, num_detectors: int, num_error_bits: int, observables: list[int] = <factory>) → None

channel_probs: list[Any]

error_transform: Any

graph: Any

num_detectors: int

num_error_bits: int

num_outputs: int

observables: list[int]

tensorcircuit.zx.converter.add_dummy(b: GraphRepresentation, qubit: int, row: float | int | None = None) → int

Add a dummy boundary vertex to a qubit lane.

Parameters:

• b (GraphRepresentation) – The graph representation.

• qubit (int) – The qubit index.

• row (float | int | None, optional) – Row index, defaults to last_row + 1.

Returns:

The index of the new vertex.

Return type:

int

tensorcircuit.zx.converter.add_lane(b: GraphRepresentation, qubit: int) → int

Initialize a new qubit lane with two boundary vertices.

Parameters:

• b (GraphRepresentation) – The graph representation.

• qubit (int) – The qubit index.

Returns:

The index of the first vertex.

Return type:

int

tensorcircuit.zx.converter.build_amplitude_graph(built: GraphRepresentation, bitstring: Sequence[int]) → GraphS

Build a scalar graph representing the amplitude <bitstring|psi>.

Parameters:

• built (GraphRepresentation) – The input ZX representation.

• bitstring (Sequence[int]) – The target bitstring as a sequence of 0s and 1s.

Returns:

The scalar ZX graph.

Return type:

GraphS

tensorcircuit.zx.converter.build_sampling_graph(built: GraphRepresentation, sample_detectors: bool, pauli: Dict[int, str] | None = None) → GraphS

Build a doubled sampling graph from a ZX representation.

Parameters:

• built (GraphRepresentation) – The input ZX representation.

• sample_detectors (bool) – Whether to prepare for detector/observable sampling or measurement records.

• pauli (Optional[Dict[int, str]]) – Optional Pauli string to insert at the junction for expectation values. Dictionary mapping qubit index to Pauli operator (‘I’, ‘X’, ‘Y’, ‘Z’).

Returns:

The prepared ZX graph.

Return type:

GraphS

tensorcircuit.zx.converter.circuit_to_zx(c: AbstractCircuit, force_measure_all: bool = False) → GraphRepresentation

Convert a TensorCircuit AbstractCircuit to a ZX-calculus GraphRepresentation.

Parameters:

• c (AbstractCircuit) – The source circuit.

• force_measure_all (bool, optional) – Whether to force measurements on all qubits at the end, defaults to False.

Returns:

The ZX graph representation of the circuit.

Return type:

GraphRepresentation

tensorcircuit.zx.converter.depolarize1(b: GraphRepresentation, qubit: int, p: float) → None

Apply single-qubit depolarizing channel.

Parameters:

• b (GraphRepresentation) – The graph representation.

• qubit (int) – Qubit index.

• p (float) – Depolarizing probability.

tensorcircuit.zx.converter.depolarize2(b: GraphRepresentation, q1: int, q2: int, p: float) → None

Apply two-qubit depolarizing channel.

Parameters:

• b (GraphRepresentation) – The graph representation.

• q1 (int) – First qubit index.

• q2 (int) – Second qubit index.

• p (float) – Depolarizing probability.

tensorcircuit.zx.converter.detector(b: GraphRepresentation, rec: list[int]) → None

Add a detector vertex defined by a set of measurement outcomes.

Parameters:

• b (GraphRepresentation) – The graph representation.

• rec (list[int]) – Indices of measurement records involved in the detector.

tensorcircuit.zx.converter.ensure_lane(b: GraphRepresentation, qubit: int) → None

Ensure a qubit lane exists in the graph.

Parameters:

• b (GraphRepresentation) – The graph representation.

• qubit (int) – The qubit index.

tensorcircuit.zx.converter.h_gate(b: GraphRepresentation, qubit: int) → None

Apply Hadamard gate.

tensorcircuit.zx.converter.h_xy(b: GraphRepresentation, qubit: int) → None

Apply variant of Hadamard gate that swaps the X and Y axes (instead of X and Z).

tensorcircuit.zx.converter.h_yz(b: GraphRepresentation, qubit: int) → None

Apply variant of Hadamard gate that swaps the Y and Z axes (instead of X and Z).

tensorcircuit.zx.converter.is_pauli(matrix: Any) → str | None

Check if a matrix is a Pauli matrix.

Parameters:

matrix (np.ndarray) – The matrix to check.

Returns:

The name of the Pauli matrix (‘i’, ‘x’, ‘y’, ‘z’) or None.

Return type:

Optional[str]

tensorcircuit.zx.converter.last_edge(b: GraphRepresentation, qubit: int) → Any

Get the last edge on a qubit lane.

Parameters:

• b (GraphRepresentation) – The graph representation.

• qubit (int) – The qubit index.

Returns:

The last edge index.

Return type:

Any

tensorcircuit.zx.converter.last_row(b: GraphRepresentation, qubit: int) → float

Get the row index of the last vertex on a qubit lane.

Parameters:

• b (GraphRepresentation) – The graph representation.

• qubit (int) – The qubit index.

Returns:

Highest row index on the lane.

Return type:

float

tensorcircuit.zx.converter.m(b: GraphRepresentation, qubit: int, p: float = 0, invert: bool = False) → None

Measure qubit in Z basis with optional bit-flip error probability p.

tensorcircuit.zx.converter.mpp(b: GraphRepresentation, paulis: list[tuple[str, int]], invert: bool = False) → None

Measure a multi-Pauli product.

Identical implementation to tsim.core.instructions.mpp.

Parameters:

• b – The graph representation to modify.

• paulis – List of (pauli_type, qubit) pairs defining the Pauli product. pauli_type should be ‘X’, ‘Y’, or ‘Z’.

• invert – Whether to invert the measurement result.

tensorcircuit.zx.converter.mr(b: GraphRepresentation, qubit: int, p: float = 0, invert: bool = False) → None

Z-basis demolition measurement (optionally noisy).

tensorcircuit.zx.converter.mrx(b: GraphRepresentation, qubit: int, p: float = 0, invert: bool = False) → None

X-basis demolition measurement (optionally noisy).

tensorcircuit.zx.converter.mry(b: GraphRepresentation, qubit: int, p: float = 0, invert: bool = False) → None

Y-basis demolition measurement (optionally noisy).

tensorcircuit.zx.converter.mrz(b: GraphRepresentation, qubit: int, p: float = 0, invert: bool = False) → None

Z-basis demolition measurement (optionally noisy).

tensorcircuit.zx.converter.mx(b: GraphRepresentation, qubit: int, p: float = 0, invert: bool = False) → None

Apply X-basis measurement.

Parameters:

• b (GraphRepresentation) – The graph representation.

• qubit (int) – Qubit index.

• p (float, optional) – Error probability, defaults to 0.

• invert (bool, optional) – Whether to invert measurement result, defaults to False.

tensorcircuit.zx.converter.my(b: GraphRepresentation, qubit: int, p: float = 0, invert: bool = False) → None

Apply Y-basis measurement.

Parameters:

• b (GraphRepresentation) – The graph representation.

• qubit (int) – Qubit index.

• p (float, optional) – Error probability, defaults to 0.

• invert (bool, optional) – Whether to invert measurement result, defaults to False.

tensorcircuit.zx.converter.observable_include(b: GraphRepresentation, rec: list[int], idx: int) → None

Add an observable vertex defined by a set of measurement outcomes.

Parameters:

• b (GraphRepresentation) – The graph representation.

• rec (list[int]) – Indices of measurement records involved in the observable.

• idx (int) – Index of the observable.

tensorcircuit.zx.converter.pauli_channel_1(b: GraphRepresentation, qubit: int, px: float = 0, py: float = 0, pz: float = 0) → None

Apply single-qubit Pauli channel.

Parameters:

• b (GraphRepresentation) – The graph representation.

• qubit (int) – Qubit index.

• px (float, optional) – Probability of X error, defaults to 0.

• py (float, optional) – Probability of Y error, defaults to 0.

• pz (float, optional) – Probability of Z error, defaults to 0.

tensorcircuit.zx.converter.pauli_channel_2(b: GraphRepresentation, qubit_i: int, qubit_j: int, pix: float = 0, piy: float = 0, piz: float = 0, pxi: float = 0, pxx: float = 0, pxy: float = 0, pxz: float = 0, pyi: float = 0, pyx: float = 0, pyy: float = 0, pyz: float = 0, pzi: float = 0, pzx: float = 0, pzy: float = 0, pzz: float = 0) → None

Apply two-qubit Pauli channel.

Parameters:

• b (GraphRepresentation) – The graph representation.

• qubit_i (int) – First qubit index.

• qubit_j (int) – Second qubit index.

• pix – Probability of error IX.

…

tensorcircuit.zx.converter.prepare_graph(circuit: AbstractCircuit, *, sample_detectors: bool, force_measure_all: bool = False, pauli: Dict[int, str] | None = None, reset_scalar: bool = True) → SamplingGraph

Prepare a circuit for sampling by converting to ZX and reducing.

Parameters:

• circuit (AbstractCircuit) – The input circuit.

• sample_detectors (bool) – Whether to prepare for detector/observable sampling.

• force_measure_all (bool, optional) – Whether to force final measurements, defaults to False.

• pauli (Optional[Dict[int, str]]) – Optional Pauli string to insert at the junction.

Returns:

A SamplingGraph object containing the reduced graph and metadata.

Return type:

SamplingGraph

tensorcircuit.zx.converter.r_x(b: GraphRepresentation, qubit: int, phase: Fraction) → None

Apply R_X rotation gate with given phase (in units of π).

tensorcircuit.zx.converter.r_y(b: GraphRepresentation, qubit: int, phase: Fraction) → None

Apply R_Y rotation gate with given phase (in units of π).

tensorcircuit.zx.converter.r_z(b: GraphRepresentation, qubit: int, phase: Fraction) → None

Apply R_Z rotation gate with given phase (in units of π).

tensorcircuit.zx.converter.reset_x(b: GraphRepresentation, qubit: int) → None

Reset qubit in X basis.

Parameters:

• b (GraphRepresentation) – The graph representation.

• qubit (int) – Qubit index.

tensorcircuit.zx.converter.reset_y(b: GraphRepresentation, qubit: int) → None

Reset qubit in Y basis.

Parameters:

• b (GraphRepresentation) – The graph representation.

• qubit (int) – Qubit index.

tensorcircuit.zx.converter.reset_z(b: GraphRepresentation, qubit: int, p: float = 0) → None

Reset qubit to |0> state (Z-basis reset).

Parameters:

• b (GraphRepresentation) – The graph representation.

• qubit (int) – Qubit index.

• p (float, optional) – Error probability, defaults to 0.

tensorcircuit.zx.converter.sqrt_x(b: GraphRepresentation, qubit: int) → None

Apply SQRT_X gate.

tensorcircuit.zx.converter.sqrt_x_dag(b: GraphRepresentation, qubit: int) → None

Apply SQRT_X_DAG gate.

tensorcircuit.zx.converter.sqrt_y(b: GraphRepresentation, qubit: int) → None

Apply SQRT_Y gate.

tensorcircuit.zx.converter.sqrt_y_dag(b: GraphRepresentation, qubit: int) → None

Apply SQRT_Y_DAG gate.

tensorcircuit.zx.converter.sqrt_z(b: GraphRepresentation, qubit: int) → None

Apply SQRT_Z gate.

tensorcircuit.zx.converter.sqrt_z_dag(b: GraphRepresentation, qubit: int) → None

Apply SQRT_Z_DAG gate.

tensorcircuit.zx.converter.squash_graph(g: Any) → None

Compact the graph layout by re-assigning qubit and row indices.

Parameters:

g (Any) – The ZX graph.

tensorcircuit.zx.converter.transform_error_basis(g: Any, num_e: int | None = None) → tuple[Any, Any]

Transform error bit variables from original ‘e’ basis to a reduced ‘f’ basis.

Parameters:

• g (Any) – The ZX graph containing error variables.

• num_e (int, optional) – Total number of error bits, defaults to None.

Returns:

A tuple of (transformed_graph, basis_transformation_matrix).

Return type:

tuple[Any, Any]

tensorcircuit.zx.converter.u3(b: GraphRepresentation, qubit: int, theta: Fraction, phi: Fraction, lambda_: Fraction) → None

Apply U3 gate: U3(θ,φ,λ) = R_Z(φ)·R_Y(θ)·R_Z(λ).

tensorcircuit.zx.converter.x_error(b: GraphRepresentation, qubit: int, p: float) → None

tensorcircuit.zx.converter.x_gate(b: GraphRepresentation, qubit: int) → None

Apply Pauli X gate.

tensorcircuit.zx.converter.x_phase(b: GraphRepresentation, qubit: int, phase: Fraction) → None

Apply an X spider with the given phase to a qubit.

Parameters:

• b (GraphRepresentation) – The graph representation.

• qubit (int) – The qubit index.

• phase (Fraction) – Phase in units of π.

tensorcircuit.zx.converter.y_error(b: GraphRepresentation, qubit: int, p: float) → None

Apply Y error with probability p.

tensorcircuit.zx.converter.y_gate(b: GraphRepresentation, qubit: int) → None

Apply Pauli Y gate.

tensorcircuit.zx.converter.y_phase(b: GraphRepresentation, qubit: int, phase: Fraction) → None

Apply Y-basis spider with the given phase.

tensorcircuit.zx.converter.z_error(b: GraphRepresentation, qubit: int, p: float) → None

tensorcircuit.zx.converter.z_gate(b: GraphRepresentation, qubit: int) → None

Apply Pauli Z gate.

tensorcircuit.zx.converter.z_phase(b: GraphRepresentation, qubit: int, phase: Fraction) → None

Apply a Z spider with the given phase to a qubit.

Parameters:

• b (GraphRepresentation) – The graph representation.

• qubit (int) – The qubit index.

• phase (Fraction) – Phase in units of π.