Fix complex cast warning

qiskit_dynamics/arraylias/register_functions/linear_combo.py

@@ -36,7 +36,10 @@ def _(coeffs, mats):
 
         @alias.register_function(lib="jax", path="linear_combo")
         def _(coeffs, mats):
-            return jnp.tensordot(coeffs, mats, axes=1)
+            # real and imag broken up to avoid real/complex tensordot warning
+            return jnp.tensordot(coeffs, mats.real, axes=1) + 1j * jnp.tensordot(
+                coeffs, mats.imag, axes=1
+            )
 
         from jax.experimental.sparse import sparsify
 

qiskit_dynamics/models/operator_collections.py

@@ -510,9 +510,18 @@ def evaluate_rhs(
                 )
 
             if self._static_dissipators is None:
+                # real and imag broken up to avoid real/complex tensordot warning
+                mats = _matmul(
+                    self._dissipator_operators, _matmul(y, self._dissipator_operators_adj)
+                )
                 both_mult_contribution = _numpy_multi_dispatch(
                     dis_coefficients,
-                    _matmul(self._dissipator_operators, _matmul(y, self._dissipator_operators_adj)),
+                    mats.real,
+                    path="tensordot",
+                    axes=(-1, -3),
+                ) + 1j * _numpy_multi_dispatch(
+                    dis_coefficients,
+                    mats.imag,
                     path="tensordot",
                     axes=(-1, -3),
                 )
@@ -522,14 +531,28 @@ def evaluate_rhs(
                     axis=-3,
                 )
             else:
-                both_mult_contribution = unp.sum(
-                    _matmul(self._static_dissipators, _matmul(y, self._static_dissipators_adj)),
-                    axis=-3,
-                ) + _numpy_multi_dispatch(
-                    dis_coefficients,
-                    _matmul(self._dissipator_operators, _matmul(y, self._dissipator_operators_adj)),
-                    path="tensordot",
-                    axes=(-1, -3),
+                # real and imag broken up to avoid real/complex tensordot warning
+                mats = _matmul(
+                    self._dissipator_operators, _matmul(y, self._dissipator_operators_adj)
+                )
+                both_mult_contribution = (
+                    unp.sum(
+                        _matmul(self._static_dissipators, _matmul(y, self._static_dissipators_adj)),
+                        axis=-3,
+                    )
+                    + _numpy_multi_dispatch(
+                        dis_coefficients,
+                        mats.real,
+                        path="tensordot",
+                        axes=(-1, -3),
+                    )
+                    + 1j
+                    * _numpy_multi_dispatch(
+                        dis_coefficients,
+                        mats.imag,
+                        path="tensordot",
+                        axes=(-1, -3),
+                    )
                 )
 
             return left_mult_contribution + right_mult_contribution + both_mult_contribution

test/dynamics/solvers/test_solver_classes.py

@@ -443,17 +443,18 @@ def test_rwa_td_lindblad_model(self):
 
     def test_signals_are_None(self):
         """Test the model signals return to being None after simulation."""
-
         ham_solver = Solver(hamiltonian_operators=[self.X])
-        ham_solver.solve(signals=[1.0], t_span=[0.0, 0.01], y0=np.array([0.0, 1.0]))
+        ham_solver.solve(signals=[1.0], t_span=[0.0, 0.01], y0=np.array([0.0, 1.0], dtype=complex))
         self.assertTrue(ham_solver.model.signals is None)
 
         lindblad_solver = Solver(hamiltonian_operators=[self.X], static_dissipators=[self.X])
-        lindblad_solver.solve(signals=[1.0], t_span=[0.0, 0.01], y0=np.eye(2))
+        lindblad_solver.solve(signals=[1.0], t_span=[0.0, 0.01], y0=np.eye(2, dtype=complex))
         self.assertTrue(lindblad_solver.model.signals == (None, None))
 
         td_lindblad_solver = Solver(hamiltonian_operators=[self.X], dissipator_operators=[self.X])
-        td_lindblad_solver.solve(signals=([1.0], [1.0]), t_span=[0.0, 0.01], y0=np.eye(2))
+        td_lindblad_solver.solve(
+            signals=([1.0], [1.0]), t_span=[0.0, 0.01], y0=np.eye(2, dtype=complex)
+        )
         self.assertTrue(td_lindblad_solver.model.signals == (None, None))
 
 
@@ -719,7 +720,7 @@ def test_jit_solve(self):
         def func(a):
             yf = self.ham_solver.solve(
                 t_span=np.array([0.0, 1.0]),
-                y0=np.array([0.0, 1.0]),
+                y0=np.array([0.0, 1.0], dtype=complex),
                 signals=[Signal(lambda t: a, 5.0)],
                 method="jax_odeint",
             ).y[-1]