We Keep Coding

algorithm for largest island in a given matrix

Given a 2 X 2 matrix, return different island sizes that is possible
For example, the following matrix should return [5, 7].
1 0 0 0 1
1 1 1 1 1
0 0 0 0 0
1 1 1 1 1
This is fairly straightforward problem. I am using a boolean visited matrix of same size and traverse the matrix in DFS fashion. I have implemented it here. for some reason, I am getting output as [1]. I tried debugging but my mind stopped working now. I am missing something silly I believe.
public class IslandConnectedCell {
public static void main(String[] args) {
int[][] input = {
{1,0,0,0,1},
{1,1,1,1,1},
{0,0,0,0,0},
{1,1,0,1,1}
};
dfsIsland(input);
}
public static void dfsIsland(int[][] input) {
int rows = input.length;
int cols = input[0].length;
List<Integer> countList = new ArrayList<>();
boolean visited[][] = new boolean[rows][cols];
for (int row = 0; row < rows; row++) {
for (int col = 0; col < cols; cols++) {
if (input[row][col] == 1 && !visited[row][col]) {
int count = mark(row, col, input, visited, rows, cols, 0);
countList.add(count);
}
}
}
System.out.println(countList);
}
public static int mark(int row, int col, int[][] input, boolean[][] visited, int rows, int cols, int count) {
if (row >= rows || row < 0 || col >= cols || col < 0) {
return 0;
}
if (input[row][col] == 0 || visited[row][col]) {
return 0;
}
visited[row][col] = true;
count+=1;
for (int i = row - 1; i <= row + 1; i++) {
for (int j = col - 1; j <= col + 1; j++) {
if (i != row || j != col) {
mark(i, j, input, visited, rows, cols, count);
}
}
}
return count;
}
}

There are two errors in your code.
See comment by Mick for first error:
One obvious problem in dfsIsland() is at least that for (int col = 0; col < cols; cols++) should probably be for (int col = 0; col < cols; col++) instead (maybe even better to use the common i and j for the row/col indices).
Second error is your use of count in the mark method, most glaringly the lack of using the return value in the recursive call. Remember, Java is pass-by-value.
Hint: I suggest you remove count as a parameter.
Once you fix the errors, output will be:
[7, 2, 2]
public class IslandConnectedCell {
public static void main(String... args) {
int[][] board = { {1,0,0,0,1},
{1,1,1,1,1},
{0,0,0,0,0},
{1,1,0,1,1} };
System.out.println(new IslandConnectedCell(board).getIslandSizes());
}
private final int[][] board;
private final int rows;
private final int cols;
public IslandConnectedCell(int[][] board) {
this.board = board;
this.rows = board.length;
this.cols = board[0].length;
}
public List<Integer> getIslandSizes() {
boolean visited[][] = new boolean[this.rows][this.cols];
List<Integer> countList = new ArrayList<>();
for (int row = 0; row < this.rows; row++)
for (int col = 0; col < this.cols; col++)
if (this.board[row][col] == 1 && ! visited[row][col])
countList.add(mark(row, col, visited));
return countList;
}
private int mark(int row, int col, boolean[][] visited) {
if (row >= this.rows || row < 0 || col >= this.cols || col < 0 || this.board[row][col] == 0 || visited[row][col])
return 0;
visited[row][col] = true;
int count = 1;
for (int r = -1; r <= 1; r++)
for (int c = -1; c <= 1; c++)
if (r != 0 || c != 0)
count += mark(row + r, col + c, visited);
return count;
}
}
UPDATE
To get the desired output of [7, 4] (original question), the board would need to use horizontal wraparound, so the two small islands on the bottom line becomes a single larger island.
That is easily accomplished by modifying one line of code to wraparound the column index using the % modulus operator:
count += mark(row + r, (col + c + this.cols) % this.cols, visited);

Sudoku Generator Algorithm Malfunction

I have a project at school where I have to create a Sudoku Program. I have managed to get a Solver algorithm working but not a Generator. I spent a lot of time online to see if people had found ways to make a working one and I found a man named Mark Fredrick Graves, Jr. (https://www.youtube.com/user/mfgravesjr) who provided a very detailed code on the creation of a sudoku grid on his GitHub (https://github.com/mfgravesjr/finished-projects/tree/master/SudokuGridGenerator). However, I felt that using a one dimensional array was unnecessarily difficult in terms of finding the equations to access lines, columns and boxes. I therefore tried to translate his code into a two dimensional array but I am running into issues with some of his sorting methods and what some variables represent like "int step = (a%2==0? rowOrigin + j: colOrigin + j*9);". What I would like to know is how to translate the methods he uses from one dimensional arrays into two dimensional arrays. Below are the methods (code snippets) in question and my half attempt at translating it myself.
public int[] generateGrid(){
ArrayList<Integer> arr = new ArrayList<Integer>(9);
solvedGrid = new int[81];
for(int i = 1; i <= 9; i++) arr.add(i);
//loads all boxes with numbers 1 through 9
for(int i = 0; i < 81; i++){
if(i%9 == 0) {
Collections.shuffle(arr);
}
int perBox = ((i / 3) % 3) * 9 + ((i % 27) / 9) * 3 + (i / 27) * 27 + (i % 3);
solvedGrid[perBox] = arr.get(i%9);
}
//tracks rows and columns that have been sorted
boolean[] sorted = new boolean[81];
for(int i = 0; i < 9; i++){
boolean backtrack = false;
//0 is row, 1 is column
for(int a = 0; a<2; a++){
//every number 1-9 that is encountered is registered
boolean[] registered = new boolean[10]; //index 0 will intentionally be left empty since there are only number 1-9.
int rowOrigin = i * 9;
int colOrigin = i;
ROW_COL: for(int j = 0; j < 9; j++){
//row/column stepping - making sure numbers are only registered once and marking which cells have been sorted
int step = (a%2==0? rowOrigin + j: colOrigin + j*9);
int num = solvedGrid[step];
if(!registered[num]) {
registered[num] = true;
}else {
//if duplicate in row/column
//box and adjacent-cell swap (BAS method)
//checks for either unregistered and unsorted candidates in same box,
//or unregistered and sorted candidates in the adjacent cells
for(int y = j; y >= 0; y--){
int scan = (a%2==0? i * 9 + y: i + 9 * y);
if(solvedGrid[scan] == num){
//box stepping
for(int z = (a%2==0? (i%3 + 1) * 3: 0); z < 9; z++){
if(a%2 == 1 && z%3 <= i%3) {
continue;
}
int boxOrigin = ((scan % 9) / 3) * 3 + (scan / 27) * 27;
int boxStep = boxOrigin + (z / 3) * 9 + (z % 3);
int boxNum = solvedGrid[boxStep];
if((!sorted[scan] && !sorted[boxStep] && !registered[boxNum]) || (sorted[scan] && !registered[boxNum] && (a%2==0? boxStep%9==scan%9: boxStep/9==scan/9))){
solvedGrid[scan] = boxNum;
solvedGrid[boxStep] = num;
registered[boxNum] = true;
continue ROW_COL;
}else if(z == 8) {
//if z == 8, then break statement not reached: no candidates available
//Preferred adjacent swap (PAS)
//Swaps x for y (preference on unregistered numbers), finds occurence of y
//and swaps with z, etc. until an unregistered number has been found
int searchingNo = num;
//noting the location for the blindSwaps to prevent infinite loops.
boolean[] blindSwapIndex = new boolean[81];
//loop of size 18 to prevent infinite loops as well. Max of 18 swaps are possible.
//at the end of this loop, if continue or break statements are not reached, then
//fail-safe is executed called Advance and Backtrack Sort (ABS) which allows the
//algorithm to continue sorting the next row and column before coming back.
//Somehow, this fail-safe ensures success.
for(int q = 0; q < 18; q++){
SWAP: for(int b = 0; b <= j; b++){
int pacing = (a%2==0? rowOrigin+b: colOrigin+b*9);
if(solvedGrid[pacing] == searchingNo){
int adjacentCell = -1;
int adjacentNo = -1;
int decrement = (a%2==0? 9: 1);
for(int c = 1; c < 3 - (i % 3); c++){
adjacentCell = pacing + (a%2==0? (c + 1)*9: c + 1);
//this creates the preference for swapping with unregistered numbers
if((a%2==0 && adjacentCell >= 81)
|| (a%2==1 && adjacentCell % 9 == 0)) {
adjacentCell -= decrement;
}else {
adjacentNo = solvedGrid[adjacentCell];
if(i%3!=0
|| c!=1
|| blindSwapIndex[adjacentCell]
|| registered[adjacentNo]) {
adjacentCell -= decrement;
}
}
adjacentNo = solvedGrid[adjacentCell];
//as long as it hasn't been swapped before, swap it
if(!blindSwapIndex[adjacentCell]){
blindSwapIndex[adjacentCell] = true;
solvedGrid[pacing] = adjacentNo;
solvedGrid[adjacentCell] = searchingNo;
searchingNo = adjacentNo;
if(!registered[adjacentNo]){
registered[adjacentNo] = true;
continue ROW_COL;
}
break SWAP;
}
}
}
}
}
//begin Advance and Backtrack Sort (ABS)
backtrack = true;
break ROW_COL;
}
}
}
}
}
}
if(a%2==0) {
for(int j = 0; j < 9; j++) {
sorted[i*9+j] = true; //setting row as sorted
}
}else if(!backtrack) {
for(int j = 0; j < 9; j++) {
sorted[i+j*9] = true; //setting column as sorted
}
}else {//reseting sorted cells through to the last iteration
//backtrack = false;
for(int j = 0; j < 9; j++) {
sorted[i*9+j] = false;
}
for(int j = 0; j < 9; j++) {
sorted[(i-1)*9+j] = false;
}
for(int j = 0; j < 9; j++) {
sorted[i-1+j*9] = false;
}
for(int j = 0; j < 81; j++) {
sorted[j] = false;
}
i-=2;
}
}
}
if(!isPerfect(solvedGrid)) {
throw new RuntimeException("ERROR: Imperfect grid generated.");
}
return solvedGrid;
}
My code (unfinished)
public int[][] generateGrid(){
ArrayList<Integer> arr = new ArrayList<Integer>(9);
ArrayList<Integer> values = new ArrayList<Integer>(9);
solvedGrid = new int[9][9];
for(int i = 1 ; i <= 9; i++) {
arr.add(i);
values.add(i);
}
//Fill all boxes with number 1 to 9
for(int i = 0; i < 9; i++) {
for(int j = 0; j < 9; j++) {
if(j == 0) {
Collections.shuffle(arr);
}
solvedGrid[(i/3)*3+(j/3)][(i%3)*3+(j%3)] = arr.get(j);
}
}
//boolean[][] sorted = new boolean[9][9];
for(int i = 0; i < 9; i++) {
for(int j = 0; j < 9; j++) {
int[] rowColValues = new int[9];
rowColValues[j] = solvedGrid[0][j];
ArrayList<Integer> occurence = new ArrayList<Integer>();
for(int k = 0; k < rowColValues.length; k++) {
occurence.add((k+1), occurence.get(k+1)+1);
if(occurence.get(k+1) != 1) {
//swap with number in the box that isn't already in rowColValues
//Not sure how to do this...
//Create a method that takes the correct variables as parameters ?
break;
}
}
//Read the sudoku row then column wise and swap values that already exist in the column or row.
}
}
print2DGrid(solvedGrid);
return solvedGrid;
}

Conway Game of Life not working count neighbours method not working

The following pice of code is my code for my Game of Life. For some reason it works very strangely. The first few steps of the game are wrong, and then the whole output turns to zeroes. I don't know which method is causing the problem, but I assume it is the count neighbors method.
The board is a 2D array called GOLBoard, and the x and y coordinates are cellRow and cellCol. Alive cells are 1, and dead cells are 0s. The way I avoid out of bounds problems are by making the board 12x12, but I use only 1 to 11 for rows and columns.
<code> #Override
public int countNeighbours(int cellRow, int cellCol) {
int neighbours = 0;
for (int i = cellRow-1; i < cellRow + 2; i++) {
for (int j = cellCol - 1; j < cellCol + 2; j++) {
if (GOLBoard[i][j] == 1) {
neighbours += 1;
}
}
}
if (GOLBoard[cellRow][cellCol] == 1) {
return(neighbours-1);
} else {
return(neighbours);
}
}
#Override
public int applyRules(int cellRow, int cellCol) {
int alive = 0;
if (GOLBoard[cellRow][cellCol] == 1) {
if (countNeighbours(cellRow, cellCol) == 2 || countNeighbours(cellRow, cellCol) == 3) {
alive = 1;
} else if (countNeighbours(cellRow, cellCol) < 2 || countNeighbours(cellRow, cellCol) > 3) {
alive = 0;
}
}
if (GOLBoard[cellRow][cellCol] == 0) {
if (countNeighbours(cellRow, cellCol) == 3) {
alive = 1;
}
}
return (alive);
}
#Override
public void takeStep() {
for (int row = 1; row < 11; row++) {
for (int col = 1; col < 11; col++) {
GOLBoard[row][col] = applyRules(row, col);
}
}
}
#Override
public String toString() {
for (int row = 1; row < 11; row++) {
for (int col = 1; col < 11; col++) {
System.out.print("[" + GOLBoard[row][col] + "]");
}
System.out.println("");
}
return("");
}
} <code>

If I am recalling the rules of GOL correctly, the evolution of the board should proceed in a series of "generations", wherein the state of each cell on a "new" board is determined solely by the conditions of the corresponding cell on the "old" board. Your program is trying to continuously evolve a single board, so that changes to previously computed cells are affecting the outcome of cells yet to be computed. Your takeStep routine should be doing something more like this:
newBoard = new int[12][12];
for (int row = 1; row < 11; row++) {
for (int col = 1; col < 11; col++) {
newBoard[row][col] = applyRules(row, col);
}
}
GOLBoard = newBoard;
As far as I can tell, your countNeighbours is OK.

Diagonal win checks in Connect 4 [duplicate]

This question already has answers here:
Connect 4 check for a win algorithm
(6 answers)
Closed 6 years ago.
I am developing a Connect 4 game with 7x7 fields, horizontal and vertical checks but I dont get diagonal checks works
I can win as long as one of the tokens is not in the last column.
This are all checks:
private static int getWinningInDiagonals() {
// Top-Left -> Bottom-Right
for (int column = 0; column < 7; column++) {
int count = 0;
for (int row = 0; row < 7; row++) {
if (field[row][column] != 0 && field[row+1][column + row - 1] == field[row][column])
count++;
else
count = 1;
if (count >= 4) {
return field[row][column];
}
}
}
// Top-Right -> Bottom-Left
for (int column = 0; column < 7; column++) {
int count = 0;
for (int row = 0; row < 7; row++) {
if (field[row][column] != 0 && field[row+1][column - row + 1] == field[row][column])
count++;
else
count = 1;
if (count >= 4) {
return field[row][column];
}
}
}
return 0;
}

For one you want to move down and right at the same time every time, you also only need to go to 3, 3 as a diagonal cannot occur any farther than that without it leaving the bounds of the array.
This should work for your top left to bottom rights if I'm right in assuming that your top left is your 0,0.
Doing top right to bottom left is a matter of changing the column and row loops and changing how the offset works.
// Top-Left -> Bottom-Right
for (int column = 0; column < 4; column++) {
for (int row = 0; row < 4; row++) {
player = 0;
if (field[row][column] != 0){
player=field[row][column];
offset = 1;
}
while (player != 0){
if (field[row + offset][column + offset] == player){
offset += 1;
}else{
player = 0;
}
}
if(offset >= 4){
return field[row][column];
}
}
}

8 Queens Not Using the Backtracking Method

I am doing a version of the 8 queens problem, but not using the backtracking method. For one of the methods, I have to "score the square", basically I need to find the number of cells that would become unavailable were there to be a queen placed in the box. My problem is that I cannot get my code to return the score of the square. Is there something wrong with my for loops or something?
import java.util.ArrayList;
public class Chessboard {
private int[][] board;
public static final int QUEEN = -2;
public static final int SQUIGGLE = -1;
/**
* constructor initializes board to be of size n-by-n and containing all
* zeros
*/
public Chessboard(int n) {
board = new int[n][n];
}
/**
* returns the board
*/
public int[][] getBoard() {
return board;
}
/**
* returns SQUIGGLE if square at row, col contains SQUIGGLE returns QUEEN if
* square at row, col contains QUEEN otherwise, counts the number of squares
* that would become unavailable if the square at row, col were to receive a
* queen; this count is returned
*/
public int scoreSquare(int row, int col) {
if (board[row][col] == -1) {
return SQUIGGLE;
} else if (board[row][col] == -2) {
return QUEEN;
}
else {
int countsquare = 1;
for (int r = 0; r < board[col].length; r++) {
countsquare++;
}
for (int c = 0; c < board[row].length; c++) {
countsquare++;
}
for (int r = row + 1, c = col + 1; r < board.length
&& c < board.length; r++, c++) {
countsquare++;
}
for (int r = row + 1, c = col - 1; r < board.length && c < 0; r++, c--) {
countsquare++;
}
for (int r = row - 1, c = col + 1; r < 0 && c < board.length; r--, c++) {
countsquare++;
}
for (int r = row - 1, c = col - 1; r < 0 && c < 0; r--, c--) {
countsquare++;
}
return countsquare;
}
}

In every loop where you compare r or c to zero, you need to use >= instead of <. For example, in this loop, you need to use c >= 0 as opposed to c < 0:
for (int r = row + 1, c = col - 1; r < board.length && c < 0; r++, c--) {
This is because you are counting down to zero and so you want to keep counting while your counter is above zero.
Finally, although it shouldn't affect your results, you don't really need your first two loops as each of them should just result in adding n (which is board.length, i.e., the size of any row or column).