Welcome to Quixotic

What is Quixotic?

Quixotic is a Python library for simple-to-use, low-code quantum computing.

Features

• Easy-to-apply quantum algorithms for a number of combinatorial optimization problems using Quantum Annealing and QAOA.
• Includes out-of-the-box support for various optimization problems like maximum clique and minimum vertex cover.
• Supports custom problem formulations as QUBOs.
• Simple-to-use execution of algorithms using both local simulation on your laptop and managed quantum computers on Amazon Braket or the D-Wave LEAP service.
• Quantum is Optional: serves as a general-purpose optimization library allowing you to solve problems on your laptop using simulated annealing and exact solvers

Install

1. pip install -U pip
2. pip install quixotic

NOTE: Python version >= 3.7 is required.

Usage Example: Find Maximum Clique in a Graph

# construct or load your input graph
import networkx as nx
seed = 1967
g = nx.erdos_renyi_graph(6, p=0.5, seed=seed)
positions = nx.spring_layout(g, seed=seed)
nx.draw(g, with_labels=True, pos=positions)

# approximate a solution using QuantumAnnealer and extract results
from quixotic.core import QuantumAnnealer
qo = QuantumAnnealer(g, task='maximum_clique').execute()
nodes = qo.results()

Executing locally.

# plot nodes comprising the solution
sub = g.subgraph(nodes)
nx.draw(g, pos=positions, with_labels=True)
nx.draw(sub, pos=positions, node_color="r", edge_color="r")

How to Execute on a Quantum Computer:

By default, Quixotic uses a local solver or simulator (e.g., quantum simulator, simulated annealing), which allows you to easily run and test on your CPU-based laptop. To run on an actual managed quantum computer hosted on Amazon Braket, simply set the backend to aws and supply the device_arn and s3_folder parameters. Here is an example of using the QuantumAnnealer on a quantum computer managed by Amazon Braket:

from quixotic.core import QuantumAnnealer
qo = QuantumAnnealer(g, task='maximum_clique',
                     backend='aws',                                               # Amazon AWS as backend
                     device_arn='arn:aws:braket:::device/qpu/d-wave/DW_2000Q_6',  # D-Wave QPU
                     s3_folder = ("amazon-braket-Your-Bucket-Name", "Your-Folder-Name"))
qo.execute()  # executes algorithm on quantum hardware managed by Amazon Braket
solution_nodes = qo.results()

For the QuantumAnnealer, you can alternatively use D-Wave LEAP as a backend by specifying backend='dwave'. However, to use a managed quantum computer, you'll need to first create an account with one of the backend providers using the instructions below:

Setting up an Amazon Braket Account

1. Create an AWS account and activate Amazon Braket.
2. During the onboarding process in the previous step, you will generate an Amazon Braket bucket of the form amazon-braket-XXXXXX. Make note of this.
3. If running Quixotic on your laptop and using a remote quantum device, you'll have to configure and store your AWS credentials locally. If using a managed notebook within Amazon Braket itself, you can skip this step.
4. Set backend='aws' in addition to the device_arn and s3_folder parameters when executing either the QuantumAnnealer or the QAOA algorithm, as shown above.

Setting up a D-Wave LEAP Account

The QuantumAnnealer can also use D-Wave LEAP as a backend instead of Amazon Braket, if following the steps below. (LEAP is D-Wave's cloud-based quantum service.)

1. Create a LEAP account and make note of the API Token assigned to you.
2. After Quixotic is installed, run the command: dwave config create. Copy and paste your API token when prompted and select defaults for everything else.
3. When using QuantumAnnealer in Quixotic, set backend='dwave' and run the code below to use the default solver:

# Executing on D-Wave quantum backend
from quixotic.core import QuantumAnnealer
solution_nodes = QuantumAnnealer(g, task='maximum_clique', backend='dwave').execute().results()