dq.random.operator

operator(
    key: PRNGKeyArray,
    dims: int | tuple[int, ...],
    hermitian: bool = True,
    *,
    batch: int | tuple[int, ...] = ()
) -> QArray

Returns a random operator.

Parameters:

• key –

  A PRNG key used as the random key.

• dims –

  Hilbert space dimension of each subsystem.

• hermitian –

  Whether to return a Hermitian operator.

• batch –

  Batch shape of the returned qarray.

Returns:

• (qarray of shape (*batch, prod(dims), prod(dims))) –

  Random operator.

Examples:

>>> key = jax.random.PRNGKey(42)

Random Hermitian operator for an individual system:

>>> dq.random.operator(key, 2)
QArray: shape=(2, 2), dims=(2,), dtype=complex64, layout=dense
[[ 0.323+0.j    -0.074-0.146j]
 [-0.074+0.146j -0.844+0.j   ]]

Random non-Hermitian operator for an individual system:

>>> dq.random.operator(key, 2, hermitian=False)
QArray: shape=(2, 2), dims=(2,), dtype=complex64, layout=dense
[[ 0.323+0.061j  0.016+0.47j ]
 [-0.163+0.762j -0.844-0.116j]]

Batched random Hermitian operators for an individual system:

>>> dq.random.operator(key, 2, batch=3)
QArray: shape=(3, 2, 2), dims=(2,), dtype=complex64, layout=dense
[[[ 0.323+0.j    -0.074-0.146j]
  [-0.074+0.146j -0.844+0.j   ]]

 [[-0.609+0.j    -0.082+0.881j]
  [-0.082-0.881j  0.703+0.j   ]]

 [[-0.06 +0.j    -0.12 +0.433j]
  [-0.12 -0.433j -0.273+0.j   ]]]

Random Hermitian operator for a composite system:

>>> dq.random.operator(key, (2, 3))
QArray: shape=(6, 6), dims=(6,), dtype=complex64, layout=dense
[[ 0.323+0.j    -0.008+0.661j -0.254+0.501j  0.004+0.093j -0.581+0.114j
  -0.359+0.114j]
 [-0.008-0.661j  0.703+0.j    -0.381+0.178j -0.167-0.305j  0.245-0.401j
  -0.597-0.076j]
 [-0.254-0.501j -0.381-0.178j  0.469+0.j     0.143+0.173j  0.457+0.518j
  -0.331-0.218j]
 [ 0.004-0.093j -0.167+0.305j  0.143-0.173j  0.685+0.j     0.613+0.355j
   0.201+0.151j]
 [-0.581-0.114j  0.245+0.401j  0.457-0.518j  0.613-0.355j -0.215+0.j
   0.269-0.797j]
 [-0.359-0.114j -0.597+0.076j -0.331+0.218j  0.201-0.151j  0.269+0.797j
   0.655+0.j   ]]

Random batched Hermitian operators for a composite system:

>>> dq.random.operator(key, (2, 3), batch=(4, 5, 6, 7)).shape
(4, 5, 6, 7, 6, 6)