Fast Sudoku Solver

A generic way to solve a Sudoku puzzle is to use recursion. Here is a sample code:

// determine the board is valid or not.
bool isValid(vector<vector<char>>& board, int row , int column, char c){
     for (int i = 0; i < 9; ++i)
         if (board[row][i] == c) return false;
     for (int j = 0; j < 9; ++j)
         if (board[j][column] == c) return false
     for (int i = 0; i < 3; ++i)
         for (int j = 0; j < 3; ++j)
             if (board[row / 3 * 3 + i][column / 3 * 3 + j] == c)
                return false;
     return true;
//empty cell is represented by char '.'
bool solve(vector<vector<char>>& board){
     for (int i = 0; i < 9; ++i){
         for (int j = 0; j < 9; ++j){
             if ('.' == board[i][j]){
                for (int k = 0; k < 9; ++k){
                    if (isValid(board, i, j, '1' + k)){
                       board[i][j] = '1' + k;
                       if (solve(board)) return true;
                       else board[i][j] = '.';
                 return false;
      return true;

However, this algorithm has a bad performance on instance with 30~40 numbers that are given on the board out of 81 numbers. The Fast Sudoku Solver is using:

  1. Backtracking: back up to the preceding variable and try a different assignment for it.

  2. Forward checking: keep track of remaining legal values for unassigned variables, and terminate search when any variable has no legal values.

  3. Heuristics: choose the variable which has the fewest “legal” moves (AKA minimum remaining values heuristics). For tie-breaker among most constrained variables, choose variable with most constraints on remaining variables. Given a variable, choose the least constraining value, which is the one that rules out the fewest values in the remaining variables.

pseudo code:
A table keeps recording legal moves for each variable after each step.
0. If there is an variable has no legal moves, backtrack.
1. Select unassigned variable x using most constrained and most constraining heuristics.
2. Order legal moves of x as {x1,...,xn} by least constraining heuristics. 
3. Assign x = x1, update legal-move table, and move to next variable.

The plot of average number of search steps performed on 71*10 random generated instances with the number of empty cells on the board from 10 to 80.

Fast Sudoku Solver Performance Plot

You can check all sample puzzles I used to test from here

Source code is coming.