
Results for a 10x10, 15x15 and 20x20 chessboard were calculated. A 30x30 board was attempted, but the program stopped executing after 8 days. In this time it finished 86 of the 361 jobs.
The program outputs a single line containing the alpha, znaut, feasibility and average iteration values. By concatenating the data and copying it into Microsoft Excel tables and contour graphs could be produced. These show how the feasibility changes with alpha and znaut.
Using a 10x10 chessboard, the program ran for about 7-8 hours. The table shows the feasibility (probability of finding a solution) to the N-queen problem when inputting the respective alpha and znaut values.
A contour graph representing this information
is:

This shows that the chance of finding a solution increases as either alpha or znaut becomes larger.
The program was then run using a 15x15 chessboard, followed by a 20x20. The 15x15 ran for 2 days and the 20x20 ran for 4.5 days, with both using 9 computers. A larger chessboard requires more iterations to find a solution and so takes longer to complete. The results for these runs are:
15x15 chessboard feasibility table
and contour graph:

20x20 chessboard feasibility table
and contour graph:

The high probability area becomes visibly larger as the size of the chessboard increases. Low alpha or znaut values continued to yield low feasibility values throughout each size of chessboard.
|
|
|