2d array vs structure of an array

Question

I am new to C language, so my question may not meet high standards

can we use

struct mat{

  int a[10];

};m[10];

instead of

int mat[10][10];

whats the difference? and which is more efficient?

Answer 1

You would have m[x].a[y] which is more complicated syntax than m[x][y] but adds nothing lexically

Answer 2

You should trust the compiler with that. However, I may suggest 2 alternatives that I know works well depending on your application.

1) use int a[10][10] but make sure you access them in the proper order when looping. The compiler will "underneath" use a single array structure so acessing in order

  for( i = 0 ; i < 10 ; i++) {
     for (j = 0 ; j < 10 ; j++) {
         // do something with a[i][j]
     }
  }

vs

  for( i = 0 ; i < 10 ; i++) {
     for (j = 0 ; j < 10 ; j++) {
         // do something with a[j][i]
     }
  }

is different in terms of performance. The later is more performant.

2) The option 1 requires extra care and is counter-intuitive. I much prefer to do

   int a[100]

and do

   for( i = 0 ; i < 100 ; i++)
      inline_function(a[i]);

where the function should be declared inline and preform the thing you have to have done. If you can avoid the function, it's even better. For example, if it's a sum, having it 2d or vector doesn't change anything.

EDIT: here is a reference that explain in the details the bit about the array order: http://www.cplusplus.com/doc/tutorial/arrays/

Answer 3

No. This is a struct which contains an array.

struct mat
{
  int a[10];
};

You can define an array of struct:

struct mat m[10];

To do it in one step:

struct mat{

  int a[10];

}m[10];

Note that you have an extra semicolon before m[10] which isn't correct syntax.

To access an element in arr, you use m[i].a[j]

This is not equivalent in syntax to a 2d array like:

int mat[10][10];

which you can access using mat[i][j]