#Sudoku Solver

A simple python implementation of recursive backtracking algorithm to solve Sudoku puzzle. The program runs in less than a second to solve the puzzle. Tested with easy, difficult and all empty puzzle.

###Dependencies:

• None:
• Sodukusolver and xumpy modules are sufficient to run this program.
• (Optional) Matplotlib:
• Matplotlib is required to graphically display the solution (shown above). Without matplotlib, the program will only generate textual output.
• (Optional) Setuptools:
• Setuptools is recommended to run setup.py. Without setuptools, setup.py will still run by using distutils package, but does not support install_requires.

###Usage:

• Input:
  • The program accepts Soduku puzzle in a .csv file. Each line should have 9 numbers separated by comma. The file should contain 9 lines.
• Execution:
  • (Optional) Execute: 'python setup.py install'. It will try to install dependencies (matplotlib).
  • Solving: Run 'python sudokusolver.py </path/to/input/file.csv>'
• Output:
  • Solution is printed to the console
  • Solution is written to </path/to/input/file>_out.csv file
  • If matplotlib is available, screenshot of figure is saved to </path/to/input/file>_out.png file

###Other Info:

• Test:
• Unit tests are in tests package
• xumpy module is used to eliminate dependency on numpy