Struct QConv2D 

pub struct QConv2D { /* private fields */ }

Quantum convolutional layer operating on 2D qubit grids

Extends QConv1D to operate on 2D arrangements of qubits, applying kernels to rectangular patches of the qubit lattice.

Implementations

impl QConv2D

pub fn new( width: usize, height: usize, kernel_width: usize, kernel_height: usize, stride_x: usize, stride_y: usize, n_params_per_kernel: usize, ) -> MLResult<Self>

Create a new 2D quantum convolutional layer

Arguments

• width - Grid width (qubits in x direction)
• height - Grid height (qubits in y direction)
• kernel_width - Kernel width
• kernel_height - Kernel height
• stride_x - Stride in x direction
• stride_y - Stride in y direction
• n_params_per_kernel - Number of rotation parameters per kernel

Example

// 4×4 qubit grid, 2×2 kernel, stride (1,1), 8 parameters per kernel
let conv = QConv2D::new(4, 4, 2, 2, 1, 1, 8)?;
// Will apply kernel at 9 positions (3×3 grid of positions)
// Total: 9 × 8 = 72 parameters

pub fn kernel_positions(&self) -> Vec<(usize, usize)>

Get the 2D positions where kernels will be applied

pub fn kernel_qubits(&self, position: (usize, usize)) -> Vec<(usize, usize)>

Get the qubit coordinates for a specific kernel position Returns (x, y) coordinates in the 2D grid

pub fn coords_to_index(&self, x: usize, y: usize) -> usize

Convert 2D coordinates to 1D qubit index

pub fn index_to_coords(&self, index: usize) -> (usize, usize)

Convert 1D qubit index to 2D coordinates

pub fn n_wires(&self) -> usize

Total number of qubits in the grid

impl QConv2D

pub fn n_parameters(&self) -> usize

Get the total number of trainable parameters

Trait Implementations

impl Clone for QConv2D

fn clone(&self) -> QConv2D

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for QConv2D

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

Formats the value using the given formatter.