Sudoku Puzzles¶

This module provides algorithms to solve Sudoku puzzles, plus tools for inputting, converting and displaying various ways of writing a puzzle or its solution(s). Primarily this is accomplished with the sage.games.sudoku.Sudoku class, though the legacy top-level sage.games.sudoku.sudoku() function is also available.

AUTHORS:

Tom Boothby (2008/05/02): Exact Cover, Dancing Links algorithm
Robert Beezer (2009/05/29): Backtracking algorithm, Sudoku class

class sage.games.sudoku.Sudoku(puzzle, verify_input=True)[source]¶

Bases: SageObject

An object representing a Sudoku puzzle. Primarily the purpose is to solve the puzzle, but conversions between formats are also provided.

INPUT:

puzzle – the first argument can take one of three forms
- list - a Python list with elements of the puzzle in row-major order, where a blank entry is a zero
- matrix - a square Sage matrix over \(\ZZ\)
- string - a string where each character is an entry of the puzzle. For two-digit entries, a = 10, b = 11, etc.
verify_input – boolean (default: True); use False if you know the input is valid

EXAMPLES:

Sage

sage: a = Sudoku('5...8..49...5...3..673....115..........2.8..........187....415..3...2...49..5...3')
sage: print(a)
+-----+-----+-----+
|5    |  8  |  4 9|
|     |5    |  3  |
|  6 7|3    |    1|
+-----+-----+-----+
|1 5  |     |     |
|     |2   8|     |
|     |     |  1 8|
+-----+-----+-----+
|7    |    4|1 5  |
|  3  |    2|     |
|4 9  |  5  |    3|
+-----+-----+-----+
sage: print(next(a.solve()))
+-----+-----+-----+
|5 1 3|6 8 7|2 4 9|
|8 4 9|5 2 1|6 3 7|
|2 6 7|3 4 9|5 8 1|
+-----+-----+-----+
|1 5 8|4 6 3|9 7 2|
|9 7 4|2 1 8|3 6 5|
|3 2 6|7 9 5|4 1 8|
+-----+-----+-----+
|7 8 2|9 3 4|1 5 6|
|6 3 5|1 7 2|8 9 4|
|4 9 1|8 5 6|7 2 3|
+-----+-----+-----+

Python

>>> from sage.all import *
>>> a = Sudoku('5...8..49...5...3..673....115..........2.8..........187....415..3...2...49..5...3')
>>> print(a)
+-----+-----+-----+
|5    |  8  |  4 9|
|     |5    |  3  |
|  6 7|3    |    1|
+-----+-----+-----+
|1 5  |     |     |
|     |2   8|     |
|     |     |  1 8|
+-----+-----+-----+
|7    |    4|1 5  |
|  3  |    2|     |
|4 9  |  5  |    3|
+-----+-----+-----+
>>> print(next(a.solve()))
+-----+-----+-----+
|5 1 3|6 8 7|2 4 9|
|8 4 9|5 2 1|6 3 7|
|2 6 7|3 4 9|5 8 1|
+-----+-----+-----+
|1 5 8|4 6 3|9 7 2|
|9 7 4|2 1 8|3 6 5|
|3 2 6|7 9 5|4 1 8|
+-----+-----+-----+
|7 8 2|9 3 4|1 5 6|
|6 3 5|1 7 2|8 9 4|
|4 9 1|8 5 6|7 2 3|
+-----+-----+-----+

Sage Live

a = Sudoku('5...8..49...5...3..673....115..........2.8..........187....415..3...2...49..5...3')
print(a)
print(next(a.solve()))

backtrack()[source]¶

Return a generator which iterates through all solutions of a Sudoku puzzle.

This function is intended to be called from the solve() method when the algorithm='backtrack' option is specified. However it may be called directly as a method of an instance of a Sudoku puzzle.

At this point, this method calls backtrack_all() which constructs all of the solutions as a list. Then the present method just returns the items of the list one at a time. Once Cython supports closures and a yield statement is supported, then the contents of backtrack_all() may be subsumed into this method and the sage.games.sudoku_backtrack module can be removed.

This routine can have wildly variable performance, with a factor of 4000 observed between the fastest and slowest \(9\times 9\) examples tested. Examples designed to perform poorly for naive backtracking, will do poorly (such as d below). However, examples meant to be difficult for humans often do very well, with a factor of 5 improvement over the \(DLX\) algorithm.

Without dynamically allocating arrays in the Cython version, we have limited this function to \(16\times 16\) puzzles. Algorithmic details are in the sage.games.sudoku_backtrack module.

EXAMPLES:

This example was reported to be very difficult for human solvers. This algorithm works very fast on it, at about half the time of the DLX solver. [sudoku:escargot]

Sage

sage: g = Sudoku('1....7.9..3..2...8..96..5....53..9...1..8...26....4...3......1..4......7..7...3..')
sage: print(g)
+-----+-----+-----+
|1    |    7|  9  |
|  3  |  2  |    8|
|    9|6    |5    |
+-----+-----+-----+
|    5|3    |9    |
|  1  |  8  |    2|
|6    |    4|     |
+-----+-----+-----+
|3    |     |  1  |
|  4  |     |    7|
|    7|     |3    |
+-----+-----+-----+
sage: print(next(g.solve(algorithm='backtrack')))
+-----+-----+-----+
|1 6 2|8 5 7|4 9 3|
|5 3 4|1 2 9|6 7 8|
|7 8 9|6 4 3|5 2 1|
+-----+-----+-----+
|4 7 5|3 1 2|9 8 6|
|9 1 3|5 8 6|7 4 2|
|6 2 8|7 9 4|1 3 5|
+-----+-----+-----+
|3 5 6|4 7 8|2 1 9|
|2 4 1|9 3 5|8 6 7|
|8 9 7|2 6 1|3 5 4|
+-----+-----+-----+

Python

>>> from sage.all import *
>>> g = Sudoku('1....7.9..3..2...8..96..5....53..9...1..8...26....4...3......1..4......7..7...3..')
>>> print(g)
+-----+-----+-----+
|1    |    7|  9  |
|  3  |  2  |    8|
|    9|6    |5    |
+-----+-----+-----+
|    5|3    |9    |
|  1  |  8  |    2|
|6    |    4|     |
+-----+-----+-----+
|3    |     |  1  |
|  4  |     |    7|
|    7|     |3    |
+-----+-----+-----+
>>> print(next(g.solve(algorithm='backtrack')))
+-----+-----+-----+
|1 6 2|8 5 7|4 9 3|
|5 3 4|1 2 9|6 7 8|
|7 8 9|6 4 3|5 2 1|
+-----+-----+-----+
|4 7 5|3 1 2|9 8 6|
|9 1 3|5 8 6|7 4 2|
|6 2 8|7 9 4|1 3 5|
+-----+-----+-----+
|3 5 6|4 7 8|2 1 9|
|2 4 1|9 3 5|8 6 7|
|8 9 7|2 6 1|3 5 4|
+-----+-----+-----+

Sage Live

g = Sudoku('1....7.9..3..2...8..96..5....53..9...1..8...26....4...3......1..4......7..7...3..')
print(g)
print(next(g.solve(algorithm='backtrack')))

This example has no entries in the top row and a half, and the top row of the solution is 987654321 and therefore a backtracking approach is slow, taking about 750 times as long as the DLX solver. [sudoku:wikipedia]

Sage

sage: c = Sudoku('..............3.85..1.2.......5.7.....4...1...9.......5......73..2.1........4...9')
sage: print(c)
+-----+-----+-----+
|     |     |     |
|     |    3|  8 5|
|    1|  2  |     |
+-----+-----+-----+
|     |5   7|     |
|    4|     |1    |
|  9  |     |     |
+-----+-----+-----+
|5    |     |  7 3|
|    2|  1  |     |
|     |  4  |    9|
+-----+-----+-----+
sage: print(next(c.solve(algorithm='backtrack')))
+-----+-----+-----+
|9 8 7|6 5 4|3 2 1|
|2 4 6|1 7 3|9 8 5|
|3 5 1|9 2 8|7 4 6|
+-----+-----+-----+
|1 2 8|5 3 7|6 9 4|
|6 3 4|8 9 2|1 5 7|
|7 9 5|4 6 1|8 3 2|
+-----+-----+-----+
|5 1 9|2 8 6|4 7 3|
|4 7 2|3 1 9|5 6 8|
|8 6 3|7 4 5|2 1 9|
+-----+-----+-----+

Python

>>> from sage.all import *
>>> c = Sudoku('..............3.85..1.2.......5.7.....4...1...9.......5......73..2.1........4...9')
>>> print(c)
+-----+-----+-----+
|     |     |     |
|     |    3|  8 5|
|    1|  2  |     |
+-----+-----+-----+
|     |5   7|     |
|    4|     |1    |
|  9  |     |     |
+-----+-----+-----+
|5    |     |  7 3|
|    2|  1  |     |
|     |  4  |    9|
+-----+-----+-----+
>>> print(next(c.solve(algorithm='backtrack')))
+-----+-----+-----+
|9 8 7|6 5 4|3 2 1|
|2 4 6|1 7 3|9 8 5|
|3 5 1|9 2 8|7 4 6|
+-----+-----+-----+
|1 2 8|5 3 7|6 9 4|
|6 3 4|8 9 2|1 5 7|
|7 9 5|4 6 1|8 3 2|
+-----+-----+-----+
|5 1 9|2 8 6|4 7 3|
|4 7 2|3 1 9|5 6 8|
|8 6 3|7 4 5|2 1 9|
+-----+-----+-----+

Sage Live

c = Sudoku('..............3.85..1.2.......5.7.....4...1...9.......5......73..2.1........4...9')
print(c)
print(next(c.solve(algorithm='backtrack')))

dlx(count_only=False)[source]¶

Return a generator that iterates through all solutions of a Sudoku puzzle.

INPUT:

count_only – boolean (default: False); if set to True the generator returned as output will simply generate None for each solution, so the calling routine can count these

OUTPUT: a generator that iterates over all the solutions

This function is intended to be called from the solve() method with the algorithm='dlx' option. However it may be called directly as a method of an instance of a Sudoku puzzle if speed is important and you do not need automatic conversions on the output (or even just want to count solutions without looking at them). In this case, inputting a puzzle as a list, with verify_input=False is the fastest way to create a puzzle.

Or if only one solution is needed it can be obtained with one call to next(), while the existence of a solution can be tested by catching the StopIteration exception with a try. Calling this particular method returns solutions as lists, in row-major order. It is up to you to work with this list for your own purposes. If you want fancier formatting tools, use the solve() method, which returns a generator that creates sage.games.sudoku.Sudoku objects.

EXAMPLES:

A \(9\times 9\) known to have one solution. We get the one solution and then check to see if there are more or not.

Sage

sage: e = Sudoku('4.....8.5.3..........7......2.....6.....8.4......1.......6.3.7.5..2.....1.4......')
sage: print(next(e.dlx()))
[4, 1, 7, 3, 6, 9, 8, 2, 5, 6, 3, 2, 1, 5, 8, 9, 4, 7, 9, 5, 8, 7, 2, 4, 3, 1, 6, 8, 2, 5, 4, 3, 7, 1, 6, 9, 7, 9, 1, 5, 8, 6, 4, 3, 2, 3, 4, 6, 9, 1, 2, 7, 5, 8, 2, 8, 9, 6, 4, 3, 5, 7, 1, 5, 7, 3, 2, 9, 1, 6, 8, 4, 1, 6, 4, 8, 7, 5, 2, 9, 3]
sage: len(list(e.dlx()))
1

Python

>>> from sage.all import *
>>> e = Sudoku('4.....8.5.3..........7......2.....6.....8.4......1.......6.3.7.5..2.....1.4......')
>>> print(next(e.dlx()))
[4, 1, 7, 3, 6, 9, 8, 2, 5, 6, 3, 2, 1, 5, 8, 9, 4, 7, 9, 5, 8, 7, 2, 4, 3, 1, 6, 8, 2, 5, 4, 3, 7, 1, 6, 9, 7, 9, 1, 5, 8, 6, 4, 3, 2, 3, 4, 6, 9, 1, 2, 7, 5, 8, 2, 8, 9, 6, 4, 3, 5, 7, 1, 5, 7, 3, 2, 9, 1, 6, 8, 4, 1, 6, 4, 8, 7, 5, 2, 9, 3]
>>> len(list(e.dlx()))
1

Sage Live

e = Sudoku('4.....8.5.3..........7......2.....6.....8.4......1.......6.3.7.5..2.....1.4......')
print(next(e.dlx()))
len(list(e.dlx()))

A \(9\times 9\) puzzle with multiple solutions. Once with actual solutions, once just to count.

Sage

sage: h = Sudoku('8..6..9.5.............2.31...7318.6.24.....73...........279.1..5...8..36..3......')
sage: len(list(h.dlx()))
5
sage: len(list(h.dlx(count_only=True)))
5

Python

>>> from sage.all import *
>>> h = Sudoku('8..6..9.5.............2.31...7318.6.24.....73...........279.1..5...8..36..3......')
>>> len(list(h.dlx()))
5
>>> len(list(h.dlx(count_only=True)))
5

Sage Live

h = Sudoku('8..6..9.5.............2.31...7318.6.24.....73...........279.1..5...8..36..3......')
len(list(h.dlx()))
len(list(h.dlx(count_only=True)))

A larger puzzle, with multiple solutions, but we just get one.

Sage

sage: j = Sudoku('....a..69.3....1d.2...8....e.4....b....5..c.......7.......g...f....1.e..2.b.8..3.......4.d.....6.........f..7.g..9.a..c...5.....8..f.....1..e.79.c....b.....2...6.....g.7......84....3.d..a.5....5...7..e...ca.....3.1.......b......f....4...d..e..g.92.6..8....')
sage: print(next(j.dlx()))
[5, 15, 16, 14, 10, 13, 7, 6, 9, 2, 3, 4, 11, 8, 12, 1, 13, 3, 2, 12, 11, 16, 8, 15, 1, 6, 7, 14, 10, 4, 9, 5, 1, 10, 11, 6, 9, 4, 3, 5, 15, 8, 12, 13, 16, 7, 14, 2, 9, 8, 7, 4, 12, 2, 1, 14, 10, 5, 16, 11, 6, 3, 15, 13, 12, 16, 4, 1, 13, 14, 9, 10, 2, 7, 11, 6, 8, 15, 5, 3, 3, 14, 5, 7, 16, 11, 15, 4, 12, 13, 8, 9, 1, 2, 10, 6, 2, 6, 13, 11, 1, 8, 5, 3, 4, 15, 14, 10, 7, 9, 16, 12, 15, 9, 8, 10, 2, 6, 12, 7, 3, 16, 5, 1, 4, 14, 13, 11, 8, 11, 3, 15, 5, 10, 4, 2, 13, 1, 6, 12, 14, 16, 7, 9, 16, 12, 14, 13, 7, 15, 11, 1, 8, 9, 4, 5, 2, 6, 3, 10, 6, 2, 10, 5, 14, 12, 16, 9, 7, 11, 15, 3, 13, 1, 4, 8, 4, 7, 1, 9, 8, 3, 6, 13, 16, 14, 10, 2, 5, 12, 11, 15, 11, 5, 9, 8, 6, 7, 13, 16, 14, 3, 1, 15, 12, 10, 2, 4, 7, 13, 15, 3, 4, 1, 10, 8, 5, 12, 2, 16, 9, 11, 6, 14, 10, 1, 6, 2, 15, 5, 14, 12, 11, 4, 9, 7, 3, 13, 8, 16, 14, 4, 12, 16, 3, 9, 2, 11, 6, 10, 13, 8, 15, 5, 1, 7]

Python

>>> from sage.all import *
>>> j = Sudoku('....a..69.3....1d.2...8....e.4....b....5..c.......7.......g...f....1.e..2.b.8..3.......4.d.....6.........f..7.g..9.a..c...5.....8..f.....1..e.79.c....b.....2...6.....g.7......84....3.d..a.5....5...7..e...ca.....3.1.......b......f....4...d..e..g.92.6..8....')
>>> print(next(j.dlx()))
[5, 15, 16, 14, 10, 13, 7, 6, 9, 2, 3, 4, 11, 8, 12, 1, 13, 3, 2, 12, 11, 16, 8, 15, 1, 6, 7, 14, 10, 4, 9, 5, 1, 10, 11, 6, 9, 4, 3, 5, 15, 8, 12, 13, 16, 7, 14, 2, 9, 8, 7, 4, 12, 2, 1, 14, 10, 5, 16, 11, 6, 3, 15, 13, 12, 16, 4, 1, 13, 14, 9, 10, 2, 7, 11, 6, 8, 15, 5, 3, 3, 14, 5, 7, 16, 11, 15, 4, 12, 13, 8, 9, 1, 2, 10, 6, 2, 6, 13, 11, 1, 8, 5, 3, 4, 15, 14, 10, 7, 9, 16, 12, 15, 9, 8, 10, 2, 6, 12, 7, 3, 16, 5, 1, 4, 14, 13, 11, 8, 11, 3, 15, 5, 10, 4, 2, 13, 1, 6, 12, 14, 16, 7, 9, 16, 12, 14, 13, 7, 15, 11, 1, 8, 9, 4, 5, 2, 6, 3, 10, 6, 2, 10, 5, 14, 12, 16, 9, 7, 11, 15, 3, 13, 1, 4, 8, 4, 7, 1, 9, 8, 3, 6, 13, 16, 14, 10, 2, 5, 12, 11, 15, 11, 5, 9, 8, 6, 7, 13, 16, 14, 3, 1, 15, 12, 10, 2, 4, 7, 13, 15, 3, 4, 1, 10, 8, 5, 12, 2, 16, 9, 11, 6, 14, 10, 1, 6, 2, 15, 5, 14, 12, 11, 4, 9, 7, 3, 13, 8, 16, 14, 4, 12, 16, 3, 9, 2, 11, 6, 10, 13, 8, 15, 5, 1, 7]

Sage Live

j = Sudoku('....a..69.3....1d.2...8....e.4....b....5..c.......7.......g...f....1.e..2.b.8..3.......4.d.....6.........f..7.g..9.a..c...5.....8..f.....1..e.79.c....b.....2...6.....g.7......84....3.d..a.5....5...7..e...ca.....3.1.......b......f....4...d..e..g.92.6..8....')
print(next(j.dlx()))

The puzzle h from above, but purposely made unsolvable with addition in second entry.

Sage

sage: hbad = Sudoku('82.6..9.5.............2.31...7318.6.24.....73...........279.1..5...8..36..3......')
sage: len(list(hbad.dlx()))
0
sage: next(hbad.dlx())
Traceback (most recent call last):
...
StopIteration

Python

>>> from sage.all import *
>>> hbad = Sudoku('82.6..9.5.............2.31...7318.6.24.....73...........279.1..5...8..36..3......')
>>> len(list(hbad.dlx()))
0
>>> next(hbad.dlx())
Traceback (most recent call last):
...
StopIteration

Sage Live

hbad = Sudoku('82.6..9.5.............2.31...7318.6.24.....73...........279.1..5...8..36..3......')
len(list(hbad.dlx()))
next(hbad.dlx())

A stupidly small puzzle to test the lower limits of arbitrary sized input.

Sage

sage: s = Sudoku('.')
sage: print(next(s.solve(algorithm='dlx')))
+-+
|1|
+-+

Python

>>> from sage.all import *
>>> s = Sudoku('.')
>>> print(next(s.solve(algorithm='dlx')))
+-+
|1|
+-+

Sage Live

s = Sudoku('.')
print(next(s.solve(algorithm='dlx')))

ALGORITHM:

The DLXCPP solver finds solutions to the exact-cover problem with a “Dancing Links” backtracking algorithm. Given a \(0-1\) matrix, the solver finds a subset of the rows that sums to the all \(1\)’s vector. The columns correspond to conditions, or constraints, that must be met by a solution, while the rows correspond to some collection of choices, or decisions. A \(1\) in a row and column indicates that the choice corresponding to the row meets the condition corresponding to the column.

So here, we code the notion of a Sudoku puzzle, and the hints already present, into such a \(0-1\) matrix. Then the sage.combinat.matrices.dlxcpp.DLXCPP solver makes the choices for the blank entries.

solve(algorithm='dlx')[source]¶

Return a generator object for the solutions of a Sudoku puzzle.

INPUT:

algorithm – string (default: 'dlx'); specify choice of solution algorithm. The two possible algorithms are 'dlx' and 'backtrack'

OUTPUT:

A generator that provides all solutions, as objects of the Sudoku class.

Calling next() on the returned generator just once will find a solution, presuming it exists, otherwise it will return a StopIteration exception. The generator may be used for iteration or wrapping the generator with list() will return all of the solutions as a list. Solutions are returned as new objects of the Sudoku class, so may be printed or converted using other methods in this class.

Generally, the DLX algorithm is very fast and very consistent. The backtrack algorithm is very variable in its performance, on some occasions markedly faster than DLX but usually slower by a similar factor, with the potential to be orders of magnitude slower. See the docstrings for the dlx() and backtrack_all() methods for further discussions and examples of performance. Note that the backtrack algorithm is limited to puzzles of size \(16\times 16\) or smaller.

EXAMPLES:

This puzzle has 5 solutions, but the first one returned by each algorithm are identical.

Sage

sage: h = Sudoku('8..6..9.5.............2.31...7318.6.24.....73...........279.1..5...8..36..3......')
sage: h
+-----+-----+-----+
|8    |6    |9   5|
|     |     |     |
|     |  2  |3 1  |
+-----+-----+-----+
|    7|3 1 8|  6  |
|2 4  |     |  7 3|
|     |     |     |
+-----+-----+-----+
|    2|7 9  |1    |
|5    |  8  |  3 6|
|    3|     |     |
+-----+-----+-----+
sage: next(h.solve(algorithm='backtrack'))
+-----+-----+-----+
|8 1 4|6 3 7|9 2 5|
|3 2 5|1 4 9|6 8 7|
|7 9 6|8 2 5|3 1 4|
+-----+-----+-----+
|9 5 7|3 1 8|4 6 2|
|2 4 1|9 5 6|8 7 3|
|6 3 8|2 7 4|5 9 1|
+-----+-----+-----+
|4 6 2|7 9 3|1 5 8|
|5 7 9|4 8 1|2 3 6|
|1 8 3|5 6 2|7 4 9|
+-----+-----+-----+
sage: next(h.solve(algorithm='dlx'))
+-----+-----+-----+
|8 1 4|6 3 7|9 2 5|
|3 2 5|1 4 9|6 8 7|
|7 9 6|8 2 5|3 1 4|
+-----+-----+-----+
|9 5 7|3 1 8|4 6 2|
|2 4 1|9 5 6|8 7 3|
|6 3 8|2 7 4|5 9 1|
+-----+-----+-----+
|4 6 2|7 9 3|1 5 8|
|5 7 9|4 8 1|2 3 6|
|1 8 3|5 6 2|7 4 9|
+-----+-----+-----+

Python

>>> from sage.all import *
>>> h = Sudoku('8..6..9.5.............2.31...7318.6.24.....73...........279.1..5...8..36..3......')
>>> h
+-----+-----+-----+
|8    |6    |9   5|
|     |     |     |
|     |  2  |3 1  |
+-----+-----+-----+
|    7|3 1 8|  6  |
|2 4  |     |  7 3|
|     |     |     |
+-----+-----+-----+
|    2|7 9  |1    |
|5    |  8  |  3 6|
|    3|     |     |
+-----+-----+-----+
>>> next(h.solve(algorithm='backtrack'))
+-----+-----+-----+
|8 1 4|6 3 7|9 2 5|
|3 2 5|1 4 9|6 8 7|
|7 9 6|8 2 5|3 1 4|
+-----+-----+-----+
|9 5 7|3 1 8|4 6 2|
|2 4 1|9 5 6|8 7 3|
|6 3 8|2 7 4|5 9 1|
+-----+-----+-----+
|4 6 2|7 9 3|1 5 8|
|5 7 9|4 8 1|2 3 6|
|1 8 3|5 6 2|7 4 9|
+-----+-----+-----+
>>> next(h.solve(algorithm='dlx'))
+-----+-----+-----+
|8 1 4|6 3 7|9 2 5|
|3 2 5|1 4 9|6 8 7|
|7 9 6|8 2 5|3 1 4|
+-----+-----+-----+
|9 5 7|3 1 8|4 6 2|
|2 4 1|9 5 6|8 7 3|
|6 3 8|2 7 4|5 9 1|
+-----+-----+-----+
|4 6 2|7 9 3|1 5 8|
|5 7 9|4 8 1|2 3 6|
|1 8 3|5 6 2|7 4 9|
+-----+-----+-----+

Sage Live

h = Sudoku('8..6..9.5.............2.31...7318.6.24.....73...........279.1..5...8..36..3......')
h
next(h.solve(algorithm='backtrack'))
next(h.solve(algorithm='dlx'))

Gordon Royle maintains a list of 48072 Sudoku puzzles that each has a unique solution and exactly 17 “hints” (initially filled boxes). At this writing (May 2009) there is no known 16-hint puzzle with exactly one solution. [sudoku:royle] This puzzle is number 3000 in his database. We solve it twice.

Sage

sage: b = Sudoku('8..6..9.5.............2.31...7318.6.24.....73...........279.1..5...8..36..3......')
sage: next(b.solve(algorithm='dlx')) == next(b.solve(algorithm='backtrack'))
True

Python

>>> from sage.all import *
>>> b = Sudoku('8..6..9.5.............2.31...7318.6.24.....73...........279.1..5...8..36..3......')
>>> next(b.solve(algorithm='dlx')) == next(b.solve(algorithm='backtrack'))
True

Sage Live

b = Sudoku('8..6..9.5.............2.31...7318.6.24.....73...........279.1..5...8..36..3......')
next(b.solve(algorithm='dlx')) == next(b.solve(algorithm='backtrack'))

These are the first 10 puzzles in a list of “Top 95” puzzles, [sudoku:top95] which we use to show that the two available algorithms obtain the same solution for each.

Sage

sage: top =['4.....8.5.3..........7......2.....6.....8.4......1.......6.3.7.5..2.....1.4......',
....:       '52...6.........7.13...........4..8..6......5...........418.........3..2...87.....',
....:       '6.....8.3.4.7.................5.4.7.3..2.....1.6.......2.....5.....8.6......1....',
....:       '48.3............71.2.......7.5....6....2..8.............1.76...3.....4......5....',
....:       '....14....3....2...7..........9...3.6.1.............8.2.....1.4....5.6.....7.8...',
....:       '......52..8.4......3...9...5.1...6..2..7........3.....6...1..........7.4.......3.',
....:       '6.2.5.........3.4..........43...8....1....2........7..5..27...........81...6.....',
....:       '.524.........7.1..............8.2...3.....6...9.5.....1.6.3...........897........',
....:       '6.2.5.........4.3..........43...8....1....2........7..5..27...........81...6.....',
....:       '.923.........8.1...........1.7.4...........658.........6.5.2...4.....7.....9.....']
sage: p = [Sudoku(top[i]) for i in range(10)]
sage: verify = [next(p[i].solve(algorithm='dlx')) == next(p[i].solve(algorithm='backtrack')) for i in range(10)]
sage: verify == [True]*10
True

Python

>>> from sage.all import *
>>> top =['4.....8.5.3..........7......2.....6.....8.4......1.......6.3.7.5..2.....1.4......',
...       '52...6.........7.13...........4..8..6......5...........418.........3..2...87.....',
...       '6.....8.3.4.7.................5.4.7.3..2.....1.6.......2.....5.....8.6......1....',
...       '48.3............71.2.......7.5....6....2..8.............1.76...3.....4......5....',
...       '....14....3....2...7..........9...3.6.1.............8.2.....1.4....5.6.....7.8...',
...       '......52..8.4......3...9...5.1...6..2..7........3.....6...1..........7.4.......3.',
...       '6.2.5.........3.4..........43...8....1....2........7..5..27...........81...6.....',
...       '.524.........7.1..............8.2...3.....6...9.5.....1.6.3...........897........',
...       '6.2.5.........4.3..........43...8....1....2........7..5..27...........81...6.....',
...       '.923.........8.1...........1.7.4...........658.........6.5.2...4.....7.....9.....']
>>> p = [Sudoku(top[i]) for i in range(Integer(10))]
>>> verify = [next(p[i].solve(algorithm='dlx')) == next(p[i].solve(algorithm='backtrack')) for i in range(Integer(10))]
>>> verify == [True]*Integer(10)
True

Sage Live

top =['4.....8.5.3..........7......2.....6.....8.4......1.......6.3.7.5..2.....1.4......',
      '52...6.........7.13...........4..8..6......5...........418.........3..2...87.....',
      '6.....8.3.4.7.................5.4.7.3..2.....1.6.......2.....5.....8.6......1....',
      '48.3............71.2.......7.5....6....2..8.............1.76...3.....4......5....',
      '....14....3....2...7..........9...3.6.1.............8.2.....1.4....5.6.....7.8...',
      '......52..8.4......3...9...5.1...6..2..7........3.....6...1..........7.4.......3.',
      '6.2.5.........3.4..........43...8....1....2........7..5..27...........81...6.....',
      '.524.........7.1..............8.2...3.....6...9.5.....1.6.3...........897........',
      '6.2.5.........4.3..........43...8....1....2........7..5..27...........81...6.....',
      '.923.........8.1...........1.7.4...........658.........6.5.2...4.....7.....9.....']
p = [Sudoku(top[i]) for i in range(10)]
verify = [next(p[i].solve(algorithm='dlx')) == next(p[i].solve(algorithm='backtrack')) for i in range(10)]
verify == [True]*10

to_ascii()[source]¶

Construct an ASCII-art version of a Sudoku puzzle. This is a modified version of the ASCII version of a subdivided matrix.

EXAMPLES:

Sage

sage: s = Sudoku('.4..32....14..3.')
sage: print(s.to_ascii())
+---+---+
|  4|   |
|3 2|   |
+---+---+
|   |1 4|
|   |3  |
+---+---+
sage: s.to_ascii()
'+---+---+\n|  4|   |\n|3 2|   |\n+---+---+\n|   |1 4|\n|   |3  |\n+---+---+'

Python

>>> from sage.all import *
>>> s = Sudoku('.4..32....14..3.')
>>> print(s.to_ascii())
+---+---+
|  4|   |
|3 2|   |
+---+---+
|   |1 4|
|   |3  |
+---+---+
>>> s.to_ascii()
'+---+---+\n|  4|   |\n|3 2|   |\n+---+---+\n|   |1 4|\n|   |3  |\n+---+---+'

Sage Live

s = Sudoku('.4..32....14..3.')
print(s.to_ascii())
s.to_ascii()

to_latex()[source]¶

Create a string of \(\LaTeX\) code representing a Sudoku puzzle or solution.

EXAMPLES:

Sage

sage: s = Sudoku('.4..32....14..3.')
sage: print(s.to_latex())
\begin{array}{|*{2}{*{2}{r}|}}\hline
&4& & \\
3&2& & \\\hline
& &1&4\\
& &3& \\\hline
\end{array}

Python

>>> from sage.all import *
>>> s = Sudoku('.4..32....14..3.')
>>> print(s.to_latex())
\begin{array}{|*{2}{*{2}{r}|}}\hline
&4& & \\
3&2& & \\\hline
& &1&4\\
& &3& \\\hline
\end{array}

Sage Live

s = Sudoku('.4..32....14..3.')
print(s.to_latex())

to_list()[source]¶

Construct a list representing a Sudoku puzzle, in row-major order.

EXAMPLES:

Sage

sage: s = Sudoku('1.......2.9.4...5...6...7...5.9.3.......7.......85..4.7.....6...3...9.8...2.....1')
sage: s.to_list()
[1, 0, 0, 0, 0, 0, 0, 0, 2, 0, 9, 0, 4, 0, 0, 0, 5, 0, 0, 0, 6, 0, 0, 0, 7, 0, 0, 0, 5, 0, 9, 0, 3, 0, 0, 0, 0, 0, 0, 0, 7, 0, 0, 0, 0, 0, 0, 0, 8, 5, 0, 0, 4, 0, 7, 0, 0, 0, 0, 0, 6, 0, 0, 0, 3, 0, 0, 0, 9, 0, 8, 0, 0, 0, 2, 0, 0, 0, 0, 0, 1]

Python

>>> from sage.all import *
>>> s = Sudoku('1.......2.9.4...5...6...7...5.9.3.......7.......85..4.7.....6...3...9.8...2.....1')
>>> s.to_list()
[1, 0, 0, 0, 0, 0, 0, 0, 2, 0, 9, 0, 4, 0, 0, 0, 5, 0, 0, 0, 6, 0, 0, 0, 7, 0, 0, 0, 5, 0, 9, 0, 3, 0, 0, 0, 0, 0, 0, 0, 7, 0, 0, 0, 0, 0, 0, 0, 8, 5, 0, 0, 4, 0, 7, 0, 0, 0, 0, 0, 6, 0, 0, 0, 3, 0, 0, 0, 9, 0, 8, 0, 0, 0, 2, 0, 0, 0, 0, 0, 1]

Sage Live

s = Sudoku('1.......2.9.4...5...6...7...5.9.3.......7.......85..4.7.....6...3...9.8...2.....1')
s.to_list()

to_matrix()[source]¶

Construct a Sage matrix over \(\ZZ\) representing a Sudoku puzzle.

EXAMPLES:

Sage

sage: s = Sudoku('.4..32....14..3.')
sage: s.to_matrix()
[0 4 0 0]
[3 2 0 0]
[0 0 1 4]
[0 0 3 0]

Python

>>> from sage.all import *
>>> s = Sudoku('.4..32....14..3.')
>>> s.to_matrix()
[0 4 0 0]
[3 2 0 0]
[0 0 1 4]
[0 0 3 0]

Sage Live

s = Sudoku('.4..32....14..3.')
s.to_matrix()

to_string()[source]¶

Construct a string representing a Sudoku puzzle.

Blank entries are represented as periods, single digits are not converted and two digit entries are converted to lower-case letters where 10 = a, 11 = b, etc. This scheme limits puzzles to at most 36 symbols.

EXAMPLES:

Sage

sage: b = matrix(ZZ, 9, 9, [ [0,0,0,0,1,0,9,0,0], [8,0,0,4,0,0,0,0,0], [2,0,0,0,0,0,0,0,0], [0,7,0,0,3,0,0,0,0], [0,0,0,0,0,0,2,0,4], [0,0,0,0,0,0,0,5,8], [0,6,0,0,0,0,1,3,0], [7,0,0,2,0,0,0,0,0], [0,0,0,8,0,0,0,0,0] ])
sage: Sudoku(b).to_string()
'....1.9..8..4.....2.........7..3..........2.4.......58.6....13.7..2........8.....'

Python

>>> from sage.all import *
>>> b = matrix(ZZ, Integer(9), Integer(9), [ [Integer(0),Integer(0),Integer(0),Integer(0),Integer(1),Integer(0),Integer(9),Integer(0),Integer(0)], [Integer(8),Integer(0),Integer(0),Integer(4),Integer(0),Integer(0),Integer(0),Integer(0),Integer(0)], [Integer(2),Integer(0),Integer(0),Integer(0),Integer(0),Integer(0),Integer(0),Integer(0),Integer(0)], [Integer(0),Integer(7),Integer(0),Integer(0),Integer(3),Integer(0),Integer(0),Integer(0),Integer(0)], [Integer(0),Integer(0),Integer(0),Integer(0),Integer(0),Integer(0),Integer(2),Integer(0),Integer(4)], [Integer(0),Integer(0),Integer(0),Integer(0),Integer(0),Integer(0),Integer(0),Integer(5),Integer(8)], [Integer(0),Integer(6),Integer(0),Integer(0),Integer(0),Integer(0),Integer(1),Integer(3),Integer(0)], [Integer(7),Integer(0),Integer(0),Integer(2),Integer(0),Integer(0),Integer(0),Integer(0),Integer(0)], [Integer(0),Integer(0),Integer(0),Integer(8),Integer(0),Integer(0),Integer(0),Integer(0),Integer(0)] ])
>>> Sudoku(b).to_string()
'....1.9..8..4.....2.........7..3..........2.4.......58.6....13.7..2........8.....'

Sage Live

b = matrix(ZZ, 9, 9, [ [0,0,0,0,1,0,9,0,0], [8,0,0,4,0,0,0,0,0], [2,0,0,0,0,0,0,0,0], [0,7,0,0,3,0,0,0,0], [0,0,0,0,0,0,2,0,4], [0,0,0,0,0,0,0,5,8], [0,6,0,0,0,0,1,3,0], [7,0,0,2,0,0,0,0,0], [0,0,0,8,0,0,0,0,0] ])
Sudoku(b).to_string()

sage.games.sudoku.sudoku(m)[source]¶

Solve Sudoku puzzles described by matrices.

INPUT:

m – a square Sage matrix over \(\ZZ\), where zeros are blank entries

OUTPUT:

A Sage matrix over \(\ZZ\) containing the first solution found, otherwise None.

This function matches the behavior of the prior Sudoku solver and is included only to replicate that behavior. It could be safely deprecated, since all of its functionality is included in the Sudoku class.

EXAMPLES:

An example that was used in previous doctests.

Sage

sage: A = matrix(ZZ,9,[5,0,0, 0,8,0, 0,4,9, 0,0,0, 5,0,0, 0,3,0, 0,6,7, 3,0,0, 0,0,1, 1,5,0, 0,0,0, 0,0,0,  0,0,0, 2,0,8, 0,0,0, 0,0,0, 0,0,0, 0,1,8, 7,0,0, 0,0,4, 1,5,0, 0,3,0, 0,0,2, 0,0,0, 4,9,0, 0,5,0, 0,0,3])
sage: A
[5 0 0 0 8 0 0 4 9]
[0 0 0 5 0 0 0 3 0]
[0 6 7 3 0 0 0 0 1]
[1 5 0 0 0 0 0 0 0]
[0 0 0 2 0 8 0 0 0]
[0 0 0 0 0 0 0 1 8]
[7 0 0 0 0 4 1 5 0]
[0 3 0 0 0 2 0 0 0]
[4 9 0 0 5 0 0 0 3]
sage: sudoku(A)
[5 1 3 6 8 7 2 4 9]
[8 4 9 5 2 1 6 3 7]
[2 6 7 3 4 9 5 8 1]
[1 5 8 4 6 3 9 7 2]
[9 7 4 2 1 8 3 6 5]
[3 2 6 7 9 5 4 1 8]
[7 8 2 9 3 4 1 5 6]
[6 3 5 1 7 2 8 9 4]
[4 9 1 8 5 6 7 2 3]

Python

>>> from sage.all import *
>>> A = matrix(ZZ,Integer(9),[Integer(5),Integer(0),Integer(0), Integer(0),Integer(8),Integer(0), Integer(0),Integer(4),Integer(9), Integer(0),Integer(0),Integer(0), Integer(5),Integer(0),Integer(0), Integer(0),Integer(3),Integer(0), Integer(0),Integer(6),Integer(7), Integer(3),Integer(0),Integer(0), Integer(0),Integer(0),Integer(1), Integer(1),Integer(5),Integer(0), Integer(0),Integer(0),Integer(0), Integer(0),Integer(0),Integer(0),  Integer(0),Integer(0),Integer(0), Integer(2),Integer(0),Integer(8), Integer(0),Integer(0),Integer(0), Integer(0),Integer(0),Integer(0), Integer(0),Integer(0),Integer(0), Integer(0),Integer(1),Integer(8), Integer(7),Integer(0),Integer(0), Integer(0),Integer(0),Integer(4), Integer(1),Integer(5),Integer(0), Integer(0),Integer(3),Integer(0), Integer(0),Integer(0),Integer(2), Integer(0),Integer(0),Integer(0), Integer(4),Integer(9),Integer(0), Integer(0),Integer(5),Integer(0), Integer(0),Integer(0),Integer(3)])
>>> A
[5 0 0 0 8 0 0 4 9]
[0 0 0 5 0 0 0 3 0]
[0 6 7 3 0 0 0 0 1]
[1 5 0 0 0 0 0 0 0]
[0 0 0 2 0 8 0 0 0]
[0 0 0 0 0 0 0 1 8]
[7 0 0 0 0 4 1 5 0]
[0 3 0 0 0 2 0 0 0]
[4 9 0 0 5 0 0 0 3]
>>> sudoku(A)
[5 1 3 6 8 7 2 4 9]
[8 4 9 5 2 1 6 3 7]
[2 6 7 3 4 9 5 8 1]
[1 5 8 4 6 3 9 7 2]
[9 7 4 2 1 8 3 6 5]
[3 2 6 7 9 5 4 1 8]
[7 8 2 9 3 4 1 5 6]
[6 3 5 1 7 2 8 9 4]
[4 9 1 8 5 6 7 2 3]

Sage Live

A = matrix(ZZ,9,[5,0,0, 0,8,0, 0,4,9, 0,0,0, 5,0,0, 0,3,0, 0,6,7, 3,0,0, 0,0,1, 1,5,0, 0,0,0, 0,0,0,  0,0,0, 2,0,8, 0,0,0, 0,0,0, 0,0,0, 0,1,8, 7,0,0, 0,0,4, 1,5,0, 0,3,0, 0,0,2, 0,0,0, 4,9,0, 0,5,0, 0,0,3])
A
sudoku(A)

Using inputs that are possible with the Sudoku class, other than a matrix, will cause an error.

Sage

sage: sudoku('.4..32....14..3.')
Traceback (most recent call last):
...
ValueError: sudoku function expects puzzle to be a matrix, perhaps use the Sudoku class

Python

>>> from sage.all import *
>>> sudoku('.4..32....14..3.')
Traceback (most recent call last):
...
ValueError: sudoku function expects puzzle to be a matrix, perhaps use the Sudoku class

Sage Live

sudoku('.4..32....14..3.')