qdeepsdk 1.0.1

Our hybrid solver for QUBO problems merges classical and quantum computing, efficiently solving complex optimization tasks. It offers fast, accurate results for applications in logistics, finance, and more.

QDeep Hybrid Solver - Quadratic Unconstrained Binary Optimization Solver

QDeep Hybrid Solver is a Python library specifically developed for efficiently solving Quadratic Unconstrained Binary Optimization (QUBO) problems through advanced hybrid quantum-classical computing methods. It provides a user-friendly interface to submit QUBO matrices and retrieve structured, optimized solutions.

Mathematical Formulation

A QUBO problem is mathematically expressed as:

\min_{x \in \{0,1\}^n} \; x^T Q x

where:

• (x) is a binary vector.
• (Q) is a symmetric matrix representing the problem's coefficients.

Features

• Type-hinted implementation using Python's TypedDict
• Comprehensive input validation for matrices
• Secure authentication token management
• Structured and detailed response handling

Installation

Install the required dependencies along with the solver:

pip install numpy requests qdeepsdk

Dependencies

• numpy
• requests
• qdeepsdk

Quantum Optimization Parameters

m_budget

• Type: Integer
• Default Value: 1000
• Description: Represents the measurement budget, which determines the maximum number of quantum measurement samples executed during the optimization process. This parameter is critical for managing quantum resource allocation while balancing solution quality and computational efficiency.

num_reads

• Type: Integer
• Default Value: 10000
• Description: Specifies the number of iterations (reads) performed by the solver. This parameter plays an integral role in the convergence of the algorithm by defining the number of independent solution attempts during the optimization process.

Usage

import numpy as np
from qdeepsdk import QDeepHybridSolver

# Initialize the solver
solver = QDeepHybridSolver()

# Set the authentication token
solver.token = "your-auth-token-here"

# Configure parameters (if different from defaults)
solver.m_budget = 50000      # Measurement budget
solver.num_reads = 10000    # Number of reads

# Define a QUBO matrix
matrix = np.array([
    [1, 0],
    [0, -1]
])

# Solve the QUBO problem
try:
    response = solver.solve(matrix)
    results = response['QdeepHybridSolver']
    print("Hybrid Solver Results:", results)
except ValueError as e:
    print(f"Error: {e}")
except requests.RequestException as e:
    print(f"API Error: {e}")

Sample Matrix and Response

Sample Matrix

[[1, 0], [0, -1]]

Sample Response

{
        "QdeepHybridSolver": {
            "configuration": [0.0, 1.0],
            "energy": -1.0,
            "time": 3.6245157718658447
        }

}

Requirements

• Python 3.7+
• NumPy
• Requests

The library includes comprehensive error handling for:

• Invalid or missing authentication tokens
• Non-square matrices
• Non-2D matrices
• Non-numpy array inputs
• API connection issues