sudokumaker 1.2

Creator: bradpython12

Last updated:

0 purchases

TODO
Add to Cart

Description:

sudokumaker 1.2

Author:
Ralf Schlatterbeck <rsc@runtux.com>

Sudoku Maker is a generator for Sudoku number puzzles. It uses a genetic
algorithm internally, so it can serve as an introduction to genetic
algorithms. The generated Sudokus are usually very hard to solve – good
for getting rid of a Sudoku addiction (or maybe not).
It also includes a simple depth-first solver (with some optimizations)
for sudoku puzzles – the solver is internally needed when generating
sudoku puzzles. The included sudoku script can be called and reads a
sudoku from standard input. It outputs the solution (if any) or if
there isn’t a single solution to the given puzzle it will output several
(up to a maximum).
It internally uses my PGApy wrapper for the PGApack genetic algorithm
library (which I’m currently also maintaining).
The representation of sudoku puzzles is a simple: 9 lines with 9 numbers
in each line, e.g.,
000704800
001000200
200050000
407090000
010000080
600008900
100000050
050000706
000069002
The numbers 1-9 represent the given numbers of the puzzle while the
zeros represent the empty tiles. A solved puzzle simply contains no
zeros. I’ve adopted the file extension .sud for this format.
The puzzle in the example was created by sudokumaker with the option
-r 42 which sets the random seed to 42 (also the default if no
-r option is given), so the puzzle should be reproduceable with this
option (on a 64bit architecture). The example above when rendered with
sudoku_as_tex and compiled with LaTeX will look something like the
following:

There are some variants of sudoku puzzles supported. The first variant
adds the diagonals (so in each of the two diagonals the numbers 1-9 must
be present), this variant can be requested with the --diagonal
option. A printed example (which was again generated with the random seed
42 but now with the --diagonal option) looks as follows:

For the diagonal variant I’ve adopted the extension .sudd – note
that the normal sudoku and the diagonal-constrained sudokus are not
compatible, if interpreted as a normal sudoku, a diagonal-constrained
sudoku would have multiple solutions.
The second variant requires that in each quadrant there are 9 distinct
colors, the same color is always at the same position in each quadrant.
The numbers 1-9 must be present on each of the colors.
For this variant I’ve adopted the extension .sudc, as with
diagonal-constrained sudokus, color-constrained sudokus would have
multiple solutions if interpreted as a normal sudoku. This variant can
be requested with the --colorconstrained option. An example looks as
follows:

Sudoku puzzles can be pretty-printed as LaTeX using the included
sudoku_as_tex program. This currently supports printing the
diagonals in yellow if the --diagonal option is given.
Color-constrained sudokus can be printed with the --colorconstrained
option. The following color map applies:


color
letter



red
r

pink
p

violet
v

grey
g

orange
o

yellow
y

light green
l

dark green
d

blue
b



If these letters are used in the color constrained sudoku, the table
above applies. Other letters can be used, too, but the color assignment
for printing will be arbitrary in that case.
A third variant, sometimes called kikagaku has irregular colored
shapes instead of the 3X3 blocks. These can currently only be printed
using the --kikagaku option to sudoku_as_tex but not be
generated using sudokumaker. The format I’m using for these is a
modification of the .sud format. It contains the numbers in the same
format as in .sud followed by lines with letters where each letter
represents a unique color. Of course each letter has to appear exactly 9
times. I’ve not included a puzzle because I currently cannot generated
them automatically. An empty puzzle in that format would look as follows:
000000000
000000000
000000000
000000000
000000000
000000000
000000000
000000000
000000000
rrrrvvvvg
brbrrvggg
brbrovvvg
bbbbovggg
blooooogp
llldopppp
ldddoypyp
llldyypyp
lddddyyyy
The same color map as for color-constrained sudoku puzzle applies.
This would be rendered as follows:

For the genetic algorithm library, my python wrapper PGApy of the
parallel genetic algorithm library (PGApack) is needed. There should
be Windows support for PGApy but I haven’t tested the latest changes on
Windows.
Version 1.2: Parallel version

If installed with a parallel (MPI) version of pgapy the evaluation can
be parallelized
Caching now done in pre_eval and endofgen: pre_eval looks
for hits in the cache and prevents that these individuals get
evaluated again. The endofgen method fills the cache.
Default for random seed is 42
Get rid of --verbose option

Version 1.1: Add licensing information

Add LICENSE file
insert license headers into python files where not yet present

Version 1.0: Mark as stable, Python3
Now marked as Development Status Production/Stable

Python3
Stable
Remove SF logo
Documentation update, pictures
Add rendered examples

Version 0.4: Packaging fixes
Fix package name again, stick to sudokumaker to avoid name collisions.

Package namespace is now sudokumaker again
Sudokumaker depends on rsclib.sourceforge.net

Version 0.3: Color, Diagonal
Now color constrained and diagonality constrained sudokus are supported.

Diagonality constrained sudokus must have the digits 1-9 in the
diagonals, too. When printing, diagonals are printed in yellow.
Color constrained sudokus have 9 additional colors which are in the
same position in every quadrant. These must have the digits 1-9, too.
When printing we chose 9 different light colors.

Version 0.2: README update
The README (and the SF homepage which is generated from it) had wrong
link to the project. Also the python package index didn’t accept one of
my classifier. Grmpf.

Fix project link in README (SF Logo)
Remove one classifier not accepted by pypi

Version 0.1: Initial Release
Sudoku Maker is a generator for Sudoku number puzzles. It uses a genetic
algorithm internally.

First Release after a long silent development

License

For personal and professional use. You cannot resell or redistribute these repositories in their original state.

Files In This Product:

Customer Reviews

There are no reviews.