qics 0.1.0

Conic solver for quantum information theory

QICS: Quantum Information Conic Solver

QICS is a primal-dual interior point solver fully implemented in Python, and is specialized towards problems arising in quantum information theory. QICS solves conic programs of the form

$$ \min_{x \in \mathbb{R}^n} \quad c^\top x \quad \text{s.t.} \quad b - Ax = 0, \ h - Gx \in \mathcal{K}, $$

where $c \in \mathbb{R}^n$, $b \in \mathbb{R}^p$, $h \in \mathbb{R}^q$, $A \in \mathbb{R}^{p \times n}$, $G \in \mathbb{R}^{q \times n}$, and $\mathcal{K} \subset \mathbb{R}^{q}$ is a Cartesian product of convex cones. Some notable cones that QICS supports include

Cone	QICS class	Description
Positive semidefinite	qics.cones.PosSemidefinite	$\{ X \in \mathbb{H}^n : X \succeq 0 \}$
Quantum entropy	qics.cones.QuantEntr	$\text{cl}\{ (t, u, X) \in \mathbb{R} \times \mathbb{R}_{++} \times \mathbb{H}^n_{++} : t \geq -u S(u^{-1} X) \}$
Quantum relative entropy	qics.cones.QuantRelEntr	$\text{cl}{ (t, X, Y) \in \mathbb{R} \times \mathbb{H}^n_{++} \times \mathbb{H}^n_{++} : t \geq S(X | Y) }$
Quantum conditional entropy	qics.cones.QuantCondEntr	$\text{cl}\{ (t, X) \in \mathbb{R} \times \mathbb{H}^{n}_{++} : t \geq -S(X) + S(\text{tr}_i(X)) \}$
Quantum key distribution	qics.cones.QuantKeyDist	$\text{cl}\{ (t, X) \in \mathbb{R} \times \mathbb{H}^n_{++} : t \geq -S(\mathcal{G}(X)) + S(\mathcal{Z}(\mathcal{G}(X))) \}$
Operator perspective epigraph	qics.cones.OpPerspecTr	$\text{cl}\{ (T, X, Y) \in \mathbb{H}^n \times \mathbb{H}^n_{++} \times \mathbb{H}^n_{++} : T \succeq P_g(X, Y) \}$

where $S(X)=-\text{tr}[X\log(X)]$ is the quantum entropy, $S(X | Y)=\text{tr}[X\log(X) - X\log(Y)]$ is the quantum relative entropy, and $P_g(X, Y)=X^{1/2} g(X^{-1/2} Y X^{-1/2}) X^{1/2}$ is the non-commutative or operator perspective.

A full list of cones which we support can be found in our documentation.

Features

• Efficient quantum relative entropy programming

  We support optimizing over the quantum relative entropy cone, as well as related cones including the quantum conditional entropy cone, as well as slices of the quantum relative entropy cone that arise when solving quantum key rates of quantum cryptographic protocols. Numerical results show that QICS solves problems much faster than existing quantum relative entropy programming solvers, such as Hypatia, DDS, and CVXQUAD.

• Efficient semidefinite programming

  We implement an efficient semidefinite programming solver which utilizes state-of-the-art techniques for symmetric cone programming, including using Nesterov-Todd scalings and exploiting sparsity in the problem structure. Numerical results show that QICS has comparable performance to state-of-the-art semidefinite programming software, such as MOSEK, SDPA, SDPT3 and SeDuMi.

• Complex-valued matrices

  Users can specify whether cones involving variables which are symmetric matrices, such as the positive semidefinite cone or quantum relative entropy cone, involve real-valued or complex-valued (i.e., Hermitian) matrix variables. Support for Hermitian matrices is embedded directly in the definition of the cone, which can be more computationally efficient than lifting into the real-valued symmetric cone.

Installation

QICS is currently supported for Python 3.8 or later, and can be directly installed from pip by calling

pip install qics

Documentation

The full documentation of the code can be found here. Technical details about our implementation can be found in our paper.

Example

We illustrate how to use QICS with a quick example below. Additional details describing this example can be found here.

 import numpy
 import qics

 # Define conic program we want to solve
 c = numpy.array([[1., 0., 0., 0., 0., 0., 0., 0., 0.]]).T
 A = numpy.array([
     [0., 1., 0., 0., 0., 0., 0., 0., 0.],
     [0., 0., 1., 1., 0., 0., 0., 0., 0.],
     [0., 0., 0., 0., 1., 0., 0., 0., 0.],
     [0., 0., 0., 0., 0., 1., 0., 0., 0.],
     [0., 0., 0., 0., 0., 0., 0., 0., 1.]
 ])
 b = numpy.array([[2., 2., 2., 1., 1.]]).T
 cones = [qics.cones.QuantRelEntr(2)]
 model = qics.Model(c=c, A=A, b=b, cones=cones)

 # Solve the conic program
 solver = qics.Solver(model)
 info = solver.solve()

See the documentation for additional examples for how to solve semidefinite programs and quantum relative entropy programs using QICS.