qp.labs.tcdq.build_expval_func

build_expval_func(config)[source]

Build an estimator for Pauli expectation values of a qubit IQP circuit.

Returns a pure function that estimates the expectation value of each Pauli observable specified in config.observables or passed at call time.

The returned function captures precomputed data from config (generator matrices, default samples, preprocessed observables) so that repeated evaluations with different parameters are fast.

Parameters:

config (CircuitConfig) – Full circuit description including gate structure, observables, and sampling parameters. See CircuitConfig for details on how to construct one.

Returns:

A function with signature:

expval_fn(
    gates_params,
    phase_fn_params=None,
    observables=None,
    key=None,
    n_samples=None,
    init_state_elems=None,
    init_state_amps=None,
) -> (expvals, variances)

where expvals is a real array of shape (n_observables,) and variances contains the estimated variance of each expectation-value estimator.

Return type:

Callable

Example

>>> import jax
>>> import jax.numpy as jnp
>>> from pennylane.labs.tcdq import CircuitConfig, build_expval_func, create_local_gates
>>> n_qubits = 4
>>> gates = create_local_gates(n_qubits, max_weight=2)
>>> config = CircuitConfig(
...     gates=gates,
...     n_samples=5000,
...     key=jax.random.PRNGKey(0),
...     n_qubits=n_qubits,
...     observables=[[3, 3, 0, 0], [0, 0, 3, 3]],  # ZZ on (0,1) and (2,3)
... )
>>> expval_fn = jax.jit(build_expval_func(config))
>>> params = jnp.zeros(len(gates))
>>> expvals, variances = expval_fn(params)
>>> expvals.shape
(2,)