Tensor Products and Partial Traces
Tensor products
To describe the states of multipartite quantum systems (such as two coupled qubits, a qubit coupled to an oscillator, etc.) we need to expand the Hilbert space by taking the tensor product of the state vectors for each of the system components. Similarly, the operators acting on the state vectors in the combined Hilbert space (describing the coupled system) are formed by taking the tensor product of the individual operators.
In QuantumToolbox, the function tensor (or kron) is used to accomplish this task. This function takes a collection of Ket or Operator as argument and returns a composite QuantumObject for the combined Hilbert space. The function accepts an arbitrary number of QuantumObject as argument. The type of returned QuantumObject is the same as that of the input(s).
A collection of QuantumObject:
tensor(sigmax(), sigmax(), sigmax())Quantum Object: type=Operator dims=[2, 2, 2] size=(8, 8) ishermitian=true
8×8 SparseMatrixCSC{ComplexF64, Int64} with 8 stored entries:
⋅ ⋅ ⋅ … ⋅ ⋅ 1.0+0.0im
⋅ ⋅ ⋅ ⋅ 1.0+0.0im ⋅
⋅ ⋅ ⋅ 1.0+0.0im ⋅ ⋅
⋅ ⋅ ⋅ ⋅ ⋅ ⋅
⋅ ⋅ ⋅ ⋅ ⋅ ⋅
⋅ ⋅ 1.0+0.0im … ⋅ ⋅ ⋅
⋅ 1.0+0.0im ⋅ ⋅ ⋅ ⋅
1.0+0.0im ⋅ ⋅ ⋅ ⋅ ⋅ or a Vector{QuantumObject}:
op_list = fill(sigmax(), 3)
tensor(op_list)Quantum Object: type=Operator dims=[2, 2, 2] size=(8, 8) ishermitian=true
8×8 SparseMatrixCSC{ComplexF64, Int64} with 8 stored entries:
⋅ ⋅ ⋅ … ⋅ ⋅ 1.0+0.0im
⋅ ⋅ ⋅ ⋅ 1.0+0.0im ⋅
⋅ ⋅ ⋅ 1.0+0.0im ⋅ ⋅
⋅ ⋅ ⋅ ⋅ ⋅ ⋅
⋅ ⋅ ⋅ ⋅ ⋅ ⋅
⋅ ⋅ 1.0+0.0im … ⋅ ⋅ ⋅
⋅ 1.0+0.0im ⋅ ⋅ ⋅ ⋅
1.0+0.0im ⋅ ⋅ ⋅ ⋅ ⋅ Please note that tensor(op_list) or kron(op_list) with op_list is a Vector is type-instable and can hurt performance. It is recommended to use tensor(op_list...) or kron(op_list...) instead. See the Section The Importance of Type-Stability for more details.
For example, the state vector describing two qubits in their ground states is formed by taking the tensor product of the two single-qubit ground state vectors:
tensor(basis(2, 0), basis(2, 0))Quantum Object: type=Ket dims=[2, 2] size=(4,)
4-element Vector{ComplexF64}:
1.0 + 0.0im
0.0 + 0.0im
0.0 + 0.0im
0.0 + 0.0imOne can generalize to more qubits by adding more component state vectors in the argument list to the tensor (or kron) function, as illustrated in the following example:
states = QuantumObject[
normalize(basis(2, 0) + basis(2, 1)),
normalize(basis(2, 0) + basis(2, 1)),
basis(2, 0)
]
tensor(states...)Quantum Object: type=Ket dims=[2, 2, 2] size=(8,)
8-element Vector{ComplexF64}:
0.4999999999999999 + 0.0im
0.0 + 0.0im
0.4999999999999999 + 0.0im
0.0 + 0.0im
0.4999999999999999 + 0.0im
0.0 + 0.0im
0.4999999999999999 + 0.0im
0.0 + 0.0imThis state is slightly more complicated, describing two qubits in a superposition between the up and down states, while the third qubit is in its ground state.
To construct operators that act on an extended Hilbert space of a combined system, we similarly pass a list of operators for each component system to the tensor (or kron) function. For example, to form the operator that represents the simultaneous action of the $\hat{\sigma}_x$ operator on two qubits:
tensor(sigmax(), sigmax())Quantum Object: type=Operator dims=[2, 2] size=(4, 4) ishermitian=true
4×4 SparseMatrixCSC{ComplexF64, Int64} with 4 stored entries:
⋅ ⋅ ⋅ 1.0+0.0im
⋅ ⋅ 1.0+0.0im ⋅
⋅ 1.0+0.0im ⋅ ⋅
1.0+0.0im ⋅ ⋅ ⋅ To create operators in a combined Hilbert space that only act on a single component, we take the tensor product of the operator acting on the subspace of interest, with the identity operators corresponding to the components that are to be unchanged. For example, the operator that represents $\hat{\sigma}_z$ on the first qubit in a two-qubit system, while leaving the second qubit unaffected:
tensor(sigmaz(), qeye(2))Quantum Object: type=Operator dims=[2, 2] size=(4, 4) ishermitian=true
4×4 SparseMatrixCSC{ComplexF64, Int64} with 4 stored entries:
1.0+0.0im ⋅ ⋅ ⋅
⋅ 1.0+0.0im ⋅ ⋅
⋅ ⋅ -1.0+0.0im ⋅
⋅ ⋅ ⋅ -1.0+0.0imExample: Constructing composite Hamiltonians
The tensor (or kron) function is extensively used when constructing Hamiltonians for composite systems. Here we’ll look at some simple examples.
Two coupled qubits
First, let’s consider a system of two coupled qubits. Assume that both the qubits have equal energy splitting, and that the qubits are coupled through a $\hat{\sigma}_x \otimes \hat{\sigma}_x$ interaction with strength $g = 0.05$ (in units where the bare qubit energy splitting is unity). The Hamiltonian describing this system is:
H = tensor(sigmaz(), qeye(2)) +
tensor(qeye(2), sigmaz()) +
0.05 * tensor(sigmax(), sigmax())Quantum Object: type=Operator dims=[2, 2] size=(4, 4) ishermitian=true
4×4 SparseMatrixCSC{ComplexF64, Int64} with 6 stored entries:
2.0+0.0im ⋅ ⋅ 0.05+0.0im
⋅ ⋅ 0.05+0.0im ⋅
⋅ 0.05+0.0im ⋅ ⋅
0.05+0.0im ⋅ ⋅ -2.0+0.0imThree coupled qubits
The two-qubit example is easily generalized to three coupled qubits:
H = tensor(sigmaz(), qeye(2), qeye(2)) +
tensor(qeye(2), sigmaz(), qeye(2)) +
tensor(qeye(2), qeye(2), sigmaz()) +
0.5 * tensor(sigmax(), sigmax(), qeye(2)) +
0.25 * tensor(qeye(2), sigmax(), sigmax())Quantum Object: type=Operator dims=[2, 2, 2] size=(8, 8) ishermitian=true
8×8 SparseMatrixCSC{ComplexF64, Int64} with 24 stored entries:
3.0+0.0im ⋅ ⋅ … ⋅ 0.5+0.0im ⋅
⋅ 1.0+0.0im 0.25+0.0im ⋅ ⋅ 0.5+0.0im
⋅ 0.25+0.0im 1.0+0.0im ⋅ ⋅ ⋅
0.25+0.0im ⋅ ⋅ 0.5+0.0im ⋅ ⋅
⋅ ⋅ 0.5+0.0im ⋅ ⋅ 0.25+0.0im
⋅ ⋅ ⋅ … -1.0+0.0im 0.25+0.0im ⋅
0.5+0.0im ⋅ ⋅ 0.25+0.0im -1.0+0.0im ⋅
⋅ 0.5+0.0im ⋅ ⋅ ⋅ -3.0+0.0imA two-level system coupled to a cavity: The Jaynes-Cummings model
The simplest possible quantum mechanical description for light-matter interaction is encapsulated in the Jaynes-Cummings model, which describes the coupling between a two-level atom and a single-mode electromagnetic field (a cavity mode). Denoting the energy splitting of the atom and cavity $\omega_a$ and $\omega_c$, respectively, and the atom-cavity interaction strength $g$, the Jaynes-Cummings Hamiltonian can be constructed as:
\[H = \frac{\omega_a}{2}\hat{\sigma}_z + \omega_c \hat{a}^\dagger \hat{a} + g (\hat{a}^\dagger \hat{\sigma}_- + \hat{a} \hat{\sigma}_+)\]
N = 6 # cavity fock space truncation
ωc = 1.25 # frequency of cavity
ωa = 1.0 # frequency of two-level atom
g = 0.75 # interaction strength
a = tensor(qeye(2), destroy(N)) # cavity annihilation operator
# two-level atom operators
σm = tensor(destroy(2), qeye(N))
σz = tensor(sigmaz(), qeye(N))
H = 0.5 * ωa * σz + ωc * a' * a + g * (a' * σm + a * σm')Quantum Object: type=Operator dims=[2, 6] size=(12, 12) ishermitian=true
12×12 SparseMatrixCSC{ComplexF64, Int64} with 22 stored entries:
0.5+0.0im ⋅ ⋅ … ⋅ ⋅
⋅ 1.75+0.0im ⋅ ⋅ ⋅
⋅ ⋅ 3.0+0.0im ⋅ ⋅
⋅ ⋅ ⋅ ⋅ ⋅
⋅ ⋅ ⋅ ⋅ ⋅
⋅ ⋅ ⋅ … 1.67705+0.0im ⋅
⋅ 0.75+0.0im ⋅ ⋅ ⋅
⋅ ⋅ 1.06066+0.0im ⋅ ⋅
⋅ ⋅ ⋅ ⋅ ⋅
⋅ ⋅ ⋅ ⋅ ⋅
⋅ ⋅ ⋅ … 4.5+0.0im ⋅
⋅ ⋅ ⋅ ⋅ 5.75+0.0imPartial trace
The partial trace is an operation that reduces the dimension of a Hilbert space by eliminating some degrees of freedom by averaging (tracing). In this sense it is therefore the converse of the tensor product. It is useful when one is interested in only a part of a coupled quantum system. For open quantum systems, this typically involves tracing over the environment leaving only the system of interest. In QuantumToolbox the function ptrace is used to take partial traces. ptrace takes one QuantumObject as an input, and also one argument sel, which marks the component systems that should be kept, and all the other components are traced out.
Remember that the index of Julia starts from 1, and all the elements in sel should be positive Integer. Therefore, the type of sel can be either Integer, Tuple, SVector, or Vector.
Although it supports also Vector type, it is recommended to use Tuple or SVector from StaticArrays.jl to improve performance. For a brief explanation on the impact of the type of sel, see the section The Importance of Type-Stability.
For example, the density matrix describing a single qubit obtained from a coupled two-qubit system is obtained via:
ψ = tensor(
basis(2, 0),
basis(2, 1),
normalize(basis(2, 0) + basis(2, 1))
)Quantum Object: type=Ket dims=[2, 2, 2] size=(8,)
8-element Vector{ComplexF64}:
0.0 + 0.0im
0.0 + 0.0im
0.7071067811865475 + 0.0im
0.7071067811865475 + 0.0im
0.0 + 0.0im
0.0 + 0.0im
0.0 + 0.0im
0.0 + 0.0imptrace(ψ, 1) # trace out 2nd and 3rd systemsQuantum Object: type=Operator dims=[2] size=(2, 2) ishermitian=true
2×2 Matrix{ComplexF64}:
1.0+0.0im 0.0+0.0im
0.0-0.0im 0.0+0.0imptrace(ψ, (1, 3)) # trace out 2nd systemQuantum Object: type=Operator dims=[2, 2] size=(4, 4) ishermitian=true
4×4 Matrix{ComplexF64}:
0.5+0.0im 0.5+0.0im 0.0+0.0im 0.0+0.0im
0.5-0.0im 0.5+0.0im 0.0+0.0im 0.0+0.0im
0.0-0.0im 0.0-0.0im 0.0+0.0im 0.0+0.0im
0.0-0.0im 0.0-0.0im 0.0-0.0im 0.0+0.0imNote that the partial trace always results in a Operator (density matrix), regardless of whether the composite system is a pure state (described by a Ket) or a mixed state (described by a Operator):
ψ1 = normalize(basis(2, 0) + basis(2, 1))
ψ2 = basis(2, 0)
ψT = tensor(ψ1, ψ2)Quantum Object: type=Ket dims=[2, 2] size=(4,)
4-element Vector{ComplexF64}:
0.7071067811865475 + 0.0im
0.0 + 0.0im
0.7071067811865475 + 0.0im
0.0 + 0.0imptrace(ψT, 1)Quantum Object: type=Operator dims=[2] size=(2, 2) ishermitian=true
2×2 Matrix{ComplexF64}:
0.5+0.0im 0.5+0.0im
0.5+0.0im 0.5+0.0imρT = tensor(ket2dm(ψ1), ket2dm(ψ1))Quantum Object: type=Operator dims=[2, 2] size=(4, 4) ishermitian=true
4×4 Matrix{ComplexF64}:
0.25+0.0im 0.25+0.0im 0.25+0.0im 0.25+0.0im
0.25+0.0im 0.25+0.0im 0.25+0.0im 0.25+0.0im
0.25+0.0im 0.25+0.0im 0.25+0.0im 0.25+0.0im
0.25+0.0im 0.25+0.0im 0.25+0.0im 0.25+0.0imptrace(ρT, 1)Quantum Object: type=Operator dims=[2] size=(2, 2) ishermitian=true
2×2 Matrix{ComplexF64}:
0.5+0.0im 0.5+0.0im
0.5+0.0im 0.5+0.0im