Add half‐open slicing support to QuakeValue

[SeroviICAI]

Description

Implement slice() method on QuakeValue, plus enhance its getitem to interpret Python’s slice objects as half-open intervals.
See Issue #2778.

Test

# quake_value_slice_test.py
import cudaq


def test_quakevalue_slice():
    """
    Demonstrates:
     1) 3 total qubits, 3 float parameters
     2) Subset of qubits [1..3) => qubits #1,#2
     3) Perform a hadamard on that slice before calling a sub-kernel
     4) Sub-kernel uses the last 2 float params for rotation
     5) The first float param is used for a final rotation on qubit #0
    """

    # 1) Main kernel with 3 float parameters
    main_kernel, main_thetas = cudaq.make_kernel(list[float])

    # 2) Allocate 3 qubits
    qubits = main_kernel.qalloc(3)

    # 3) We'll do a final rotation on qubit #0 with the first parameter
    #    at the end, so we'll do that last. Let's proceed with slicing #1..2
    sub_qubits = qubits[1:3]  # qubits[1], qubits[2]
    sub_thetas = main_thetas[1:3]  # the 2nd and 3rd float

    # 4) Show how slicing can do more than sub-kernel calls:
    #    Apply hadamard across sub_qubits just by referencing the slice
    main_kernel.h(sub_qubits)

    # 5) Create a sub-kernel that takes 2 qubits + 2 float parameters
    sub_kernel, sk_qubits, sk_thetas = cudaq.make_kernel(cudaq.qvector, list[float])
    # sub_kernel => simple rotation
    sub_kernel.ry(sk_thetas[0], sk_qubits[0])
    sub_kernel.ry(sk_thetas[1], sk_qubits[1])

    # 6) Now call that sub-kernel in the main kernel on sub_qubits, sub_thetas
    main_kernel.apply_call(sub_kernel, sub_qubits, sub_thetas)

    # 7) Now do the final rotation with the *first param* main_thetas[0] on qubit #0
    main_kernel.rx(main_thetas[0], qubits[0])

    # 8) Measure
    main_kernel.mz(qubits)

    print("=== Kernel IR ===")
    print(main_kernel)

    # Provide some param values: param[0] is used on qubit #0,
    # param[1], param[2] used in sub-kernel for qubits #1,#2
    param_values = [0.15, 0.3, 0.45]
    result = cudaq.sample(main_kernel, param_values, shots_count=100)

    print("\nCircuit drawing:\n")
    print(cudaq.draw(main_kernel, param_values))
    print("\nSampled results:\n", result)


if __name__ == "__main__":
    test_quakevalue_slice()

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.