Converting a 2d array loop to a 1d matrix array C++

Question

I want to make the grid and nextgrid arrays into 1d arrays. So, that I can implement per cell threading more easily in CUDA. Currently It causes a segmentation fault when passed particular dimensions from command line. E.g. if gridWidth =6 and gridHeight =8. But will work as intended on some values. E.g. if they are both passed in 6.

So, I would like to know how to correctly initialize a 1d array with pair values and then how to implement this in the nested for loop.

#include <stdio.h>
#include <stdlib.h>
#include <iostream>
#include <cstdlib>
int main (int argc, char *argv[]) 
{

using namespace std;
//delcare variables
int numIterate = atoi (argv[1]);

int gridWidth = atoi (argv[2]);

int gridHeight = atoi (argv[3]);
cout << "iterate: " << numIterate;

cout << " w: " << gridWidth;

cout << " h: " << gridHeight;

cout << endl;

  //make grid

  int grid[gridWidth][gridHeight];
  int nextgrid[gridWidth][gridHeight];
  int genCount=0;

//set all values in grid as 0
    for(int i=0;i<gridHeight;i++){
        for(int j=0;j<gridWidth;j++){
            grid[i][j]=0;

        }
    }
    //insert glider pattern
    grid[2][3]=1;
    grid[3][2]=1;
    grid[4][2]=1;
    grid[4][3]=1;
    grid[4][4]=1;

    for(int k=0; k< numIterate+1; k++){

      cout << endl;
        cout << endl;

        cout << "GENERATION " << genCount <<endl;
    for(int i=0;i<gridHeight;i++){
          cout << endl;
        for(int j=0;j<gridWidth;j++){

            int count=0;
         count+= grid[i][(j-1+gridWidth)%gridWidth] //left
         +grid[i][(j+1+gridWidth)%gridWidth]    //right
         +grid[(i-1+gridHeight)%gridHeight][j]    //up
         +grid[(i+1+gridHeight)%gridHeight][j]    //down
         +grid[(i-1+gridHeight)%gridHeight][ (j-1+gridWidth)%gridWidth]  //up-left
         +grid[(i+1+gridHeight)%gridHeight][ (j+1+gridWidth)%gridWidth]  //down-right
         +grid[(i+1+gridHeight)%gridHeight][ (j-1+gridWidth)%gridWidth]  //down-left
         +grid[(i-1+gridHeight)%gridHeight][ (j+1+gridWidth)%gridWidth] ;//up-right
      //return count;
      //cout << count;
      printf("%d", grid[i][j]);

                // Cell is lonely and dies
                if ((grid[i][j] == 1) && (count < 2)){
                    nextgrid[i][j] = 0;
                }
                // Cell dies due to over population
                else if((grid[i][j] == 1) && (count > 3)){
                    nextgrid[i][j] = 0;
                }
                // A new cell is born
                else if ((grid[i][j] == 0) && (count == 3)){
                    nextgrid[i][j] = 1;
                }
                // Remains the same
                else{
                    nextgrid[i][j] = grid[i][j];
            }
        }
    }
//assign current grid to be the next grid
  for(int i=0;i<gridHeight;i++){
        for(int j=0;j<gridWidth;j++){
      grid[i][j] = nextgrid[i][j];
    }
    }
    genCount++;
    }
}

how do you want your array to merge? first array followed by the second etc. or do you want to zip the arrays like so : `arr[0][0] -> arr[1][0] -> arr[2][0] -> arr[0][1] -> arr[1][1] -> arr[2][1]` etc. — Sirmyself, Feb 27 '18 at 20:53
Isn't it [similar](https://stackoverflow.com/questions/19913596/c-2d-array-to-1d-array)? — Yotam Salmon, Feb 27 '18 at 21:09

score 0 · Accepted Answer · answered Feb 27 '18 at 20:56

Say grid1d[z] is the 1-dimensional representation of grid[x][y] where x, y are sizes of the 2d matrix and z is the size of the 1d array (also z = x * y). Say you want to use the element corresponding to gird[x1][x2] (0 <=x1 < x,0 <= y1 < y) then you have grid1d[z1] where z1 = x1 * y + y1.

Converting a 2d array loop to a 1d matrix array C++

1 Answers1