Struct TensorNetwork 

pub struct TensorNetwork { /* private fields */ }

A tensor network representation of a quantum state

Implementations

impl TensorNetwork

pub fn new(num_qubits: usize) -> Self

Create a new tensor network representing the |0…0⟩ state

pub fn apply_gate( &mut self, gate_tensor: Tensor, qubit_index: usize, ) -> QuantRS2Result<()>

Record a single-qubit gate application.

The gate tensor (rank 2, layout [out, in]) is appended to the ordered gate list. It is later contracted into the state tensor by Self::contract_to_statevector; nothing is fabricated here.

pub fn apply_two_qubit_gate( &mut self, gate_tensor: Tensor, control_index: usize, target_index: usize, ) -> QuantRS2Result<()>

Record a two-qubit gate application.

The gate tensor (rank 4, layout [out_c, out_t, in_c, in_t]) and the [control, target] qubit pair are appended to the ordered gate list for later contraction into the state tensor.

pub fn contract_to_statevector(&self) -> QuantRS2Result<Vec<Complex64>>

Contract the entire network to produce the final state vector.

This performs a genuine tensor contraction: the state is held as a rank-N tensor (N = number of qubits, every index of dimension 2) initialized to |0…0⟩, and each recorded gate tensor is contracted into it in application order (a sequential contraction path). The resulting amplitudes are read out in the crate-standard little-endian basis (qubit q occupies bit q of the flat index), so the output matches the state-vector simulator exactly. No circuit-type heuristics or canned states are used — the actual gate matrices and qubit indices fully determine the result.

Trait Implementations

impl Clone for TensorNetwork

fn clone(&self) -> TensorNetwork

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl ContractableNetwork for TensorNetwork

fn contract_tensors( &mut self, tensor_id1: usize, tensor_id2: usize, ) -> QuantRS2Result<usize>

Genuinely contract the two referenced tensors, replacing them with their contraction over any shared bonds (or their outer product if they share none) and returning the ID of the merged tensor.

fn optimize_contraction_order(&mut self) -> QuantRS2Result<()>

Compute and cache a greedy contraction order over the current network.

This runs the real greedy path optimizer (contraction::calculate_greedy_contraction_path) and stores the result for reuse; it is no longer a no-op placeholder.

impl Debug for TensorNetwork

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.