The n-queens problem refers to the problem of placing n queens on an n*n chessboard, such that no two queens are able to attack each other.

Here is an example solution for a 6-queens problem:

Though the n-queens problem is more often studied theoretically, it has been shown to have its applications. For instance, in a system requiring deadlock prevention, solving the n-queens problem would be equivalent to finding a set of deadlock free paths [4]. Other practical applications utilizing the n-queens problem include parallel memory storage schemes, VLSI testing, and traffic control [4].

This example demonstrates how to formulate the n-queens problem as a quadratic unconstrained binary optimization (QUBO) problem, which we then solve with Leap's hybrid solvers.

To run this example:

python n_queens.py

The program prompts the user to enter the number of queens (n) to place. It should be noted that larger n will take longer to run and we do not recommend running this example on n > 200.

The solution image file ('n-queens-solution.png') will be saved in the root directory.

We formulate the n-queens problem as a generalized exact cover problem with four types of constraints:

1. Exactly one queen in each column.
2. Exactly one queen in each row.
3. At most one queen in each diagonal from top-left to bottom-right.
4. At most one queen in each diagonal from bottom-left to top-right.

Here is a brief overview of the code:

• Represent each constraint with a unique number (ID).
• Represent each position on the chessboard with a subset of constraint IDs.
• Form a binary quadratic model (BQM) using these subsets of constraints.
• Run the problem (solve the BQM).
• Validate the solution.
• Plot the solution on a chessboard and save the solution image file.

Some notes to consider:

• Since there is exactly one queen on each row and column, we utilize a generalized version of the exact cover algorithm (specified in [1]) to handle the row and column constraints. This code can be found in exact_cover.py. The diagonal constraints are handled separately in n_queens.py.

• Each position on the chessboard (each subset of constraint IDs) becomes a variable in the BQM. That means that there are n**2 variables.

[1] Andrew Lucas, "Ising formulations of many NP problems", doi:10.3389/fphy.2014.00005

[2] Thijs Metsch, "Dancing links, algorithm X and the n-queens puzzle", http://www.nohuddleoffense.de/2019/01/20/dancing-links-algorithm-x-and-the-n-queens-puzzle/

[3] Wikipedia contributors, "Eight queens puzzle" Wikipedia, The Free Encyclopedia, https://en.wikipedia.org/wiki/Eight_queens_puzzle

[4] Jordan Bell & Brett Stevens, "A survey of known results and research areas for n-queens", doi.org/10.1016/j.disc.2007.12.043

Released under the Apache License 2.0. See LICENSE file.