Game of Life in Java

Question

I wrote John Conway's Game of Life in Java:

class GameOfLife {
 static int countSurrounding(int[][] board, int a, int b) {
 int count = 0;
 int[][] surrounding = {{a - 1, b - 1},
 {a - 1, b },
 {a - 1, b + 1},
 {a , b - 1},
 {a , b + 1},
 {a + 1, b - 1},
 {a + 1, b },
 {a + 1, b + 1}};
 for (int i[]: surrounding) {
 try {
 if (board[i[0]][i[1]] == 1) {
 count++;
 }
 }
 catch (ArrayIndexOutOfBoundsException e) {}
 }
 return count;
 }
 static void printBoard(int[][] board) {
 for (int[] i: board) {
 for (int j: i) {
 if (j == 1) {
 System.out.print("#");
 }
 else {
 System.out.print(".");
 }
 }
 System.out.println();
 }
 System.out.println();
 }
 public static void main(String[] args) {
 int[][] nextBoard = {{0, 0, 0, 0, 0, 0, 0, 0, 0},
 {0, 0, 0, 1, 0, 0, 0, 0, 0},
 {0, 1, 0, 1, 0, 0, 0, 0, 0},
 {0, 0, 1, 1, 0, 0, 0, 0, 0},
 {0, 0, 0, 0, 0, 0, 0, 0, 0},
 {0, 0, 0, 0, 0, 0, 0, 0, 0},
 {0, 0, 0, 0, 0, 0, 0, 0, 0},
 {0, 0, 0, 0, 0, 0, 0, 0, 0},
 {0, 0, 0, 0, 0, 0, 0, 0, 0}};
 int[][] board = new int[nextBoard.length][nextBoard[0].length];
 for (int gen = 0; gen < 25; gen++) {
 for (int i = 0; i < nextBoard.length; i++) {
 for (int j = 0; j < nextBoard[i].length; j++) {
 board[i][j] = nextBoard[i][j];
 }
 }
 printBoard(board);
 for (int i = 0; i < board.length; i++) {
 for (int j = 0; j < board[i].length; j++) {
 if (board[i][j] == 1 && !(countSurrounding(board, i, j) == 2 || countSurrounding(board, i, j) == 3)) {
 nextBoard[i][j] = 0;
 }
 else if (board[i][j] == 0 && countSurrounding(board, i, j) == 3) {
 nextBoard[i][j] = 1;
 }
 }
 }
 }
 }
}

How can I improve and optimize it?

Answer 1

A few suggestions:

• Each cell in the Game of Life has exactly two states: dead or alive. Unless you plan to implement -1 as undead, I'd suggest you use booleans instead of integers to indicate the state.

• The int i[] is a bit odd; i is normally used for indices, not for whole rows.

• In countSurrounding you are using Exceptions to count the surrounding living cells. This is wasteful and unnecessary; we can determine the correct bounds ourself.

  Given an index i in an array row, we can get the lower index min and the upper index max by

  int minRow = i == 0 ? 0 : i - 1;
  int maxRow = i == (row.length - 1) ? (row.length - 1) : i + 1;
  

  The same thing for columns (j index in an array column):

  int minCol = j == 0 ? 0 : j - 1;
  int maxCol = j == (column.length - 1) ? (column.length - 1) : j + 1;
  

  This itself assumes that i and j always are inside the bounds. We can then loop over the surrounding cells like

  for (int row_idx = minRow; row_idx <= maxRow; row_idx++) {
   for (int idx = minCol; idx <= maxCol; idx++) {
   if (board[row_idx][idx] && !(row_idx == i && idx == j)) { // assumes board is implemented with booleans
   count++
   }
   }
  }
  

  While this is longer, this doesn't abuse exceptions for something totally non-exceptional.

• The printBoard is fine, except for my first two criticisms. The if/else can be condensed to a ternary; I'd find this easier to read:

  for (boolean row[] : board) {
   for (boolean cell : row) {
   System.out.print( cell ? "#" : "." );
   }
   System.out.println();
  }
  

• In main, you initialize nextBoard with the initial board, then copy it over into board. This is slightly useless. I'd propose to refactor your main into a method boolean[][] nextGeneration(boolean[][] board). Your main would then look like

  int[][] initialBoard = ...;
  public static void main(String[] args) {
   int[][] board = new int[nextBoard.length][nextBoard[0].length];
   // copy over the initial board
   for (int i = 0; i < initialBoard.length; i++) {
   for (int j = 0; j < initalBoard[i].length; j++) {
   board[i][j] = initialBoard[i][j];
   }
   }
   // loop over 25 generations
   printBoard(board);
   for (int gen = 0; gen < 25; gen++) {
   System.out.println("\nGeneration " + gen);
   board = nextGeneration(board);
   printBoard(board);
   }
  }
  

• Now something is becoming obvious: We have collected a few static methods that operate on our board array. This is a sign that we should create a seperate class:

  Board
  ========
  - board: boolean[][]
  --------
  - countSurrounding(i: int, j: int): int
  + toString(): String
  + nextGeneration(): Board
  

  The class could be easily extended to have a static default board, or a constructor that creates a random board.

• Each "Game of Life" can be described by a ruleset like 23/3. The digits before the slash determine when a living cell stays alive, the digits after the slash determine when a new cell is born. Generalizing nextGeneration to be able to play with all possible rules is a nice exercise.

  private boolean nextCellState(int i, int j) {
   int count = countSurrounding(i, j);
   int rules[] = board[i][j] ? stayAlive : birthNew;
   for (int rule : rules) {
   if (rule == count) {
   return true;
   }
   }
   return false;
  }
  

  then

  public Board nextGeneration() {
   boolean next[][] = new boolean[board.length][board[0].length];
   int rows = board.length;
   int cols = board.[0].length;
   for (int i = 0; i < rows; i++) {
   for (int j = 0; j < cols; j++) {
   next[i][j] = nextCellState(i, j);
   }
   }
   return new Board(next);
  }
  

• Christian Hujer: You had my +1 at "undead". Great example of excessive value spaces.