Address cvxpy vec order warnings

cvxpy's vec() function will change the order in which it unravels
matrices in the future so the order needs to be passed explicitly in
order to keep the current behavior.

qiskit_experiments/library/tomography/fitters/cvxpy_lstsq.py

@@ -300,8 +300,8 @@ def cvxpy_linear_lstsq(
             idx = 0
             for i in range(num_circ_components):
                 for j in range(num_tomo_components):
-                    model = bms_r[idx] @ cvxpy.vec(rhos_r[idx]) - bms_i[idx] @ cvxpy.vec(
-                        rhos_i[idx]
+                    model = bms_r[idx] @ cvxpy.vec(rhos_r[idx], order="F") - bms_i[idx] @ cvxpy.vec(
+                        rhos_i[idx], order="F"
                     )
                     data = probability_data[i, j]
                     args.append(model - data)

qiskit_experiments/library/tomography/fitters/cvxpy_utils.py

@@ -234,8 +234,8 @@ def partial_trace_constaint(
     ptr = partial_trace_super(input_dim, output_dim)
     vec_cons = np.ravel(constraint, order="F")
     return [
-        ptr @ cvxpy.vec(mat_r) == vec_cons.real.round(12),
-        ptr @ cvxpy.vec(mat_i) == vec_cons.imag.round(12),
+        ptr @ cvxpy.vec(mat_r, order="F") == vec_cons.real.round(12),
+        ptr @ cvxpy.vec(mat_i, order="F") == vec_cons.imag.round(12),
     ]
 
 
@@ -274,7 +274,7 @@ def trace_preserving_constaint(
     output_dim = sdim // input_dim
 
     ptr = partial_trace_super(input_dim, output_dim)
-    cons = [ptr @ cvxpy.vec(arg_r) == np.identity(input_dim).ravel()]
+    cons = [ptr @ cvxpy.vec(arg_r, order="F") == np.identity(input_dim).ravel()]
 
     if hermitian:
         return cons
@@ -286,7 +286,7 @@ def trace_preserving_constaint(
         arg_i = mat_i
     else:
         raise TypeError("Input must be a cvxpy variable or list of variables")
-    cons.append(ptr @ cvxpy.vec(arg_i) == np.zeros(input_dim**2))
+    cons.append(ptr @ cvxpy.vec(arg_i, order="F") == np.zeros(input_dim**2))
     return cons