I want to allocate memory for a 3D array...

[momotaro]

I want to allocate memoru for a 3D array...and this is my code:

int*** build_3D_arr(int n, int ***arr){
arr = (int ***) calloc (sizeof(int**));
for(int i = 0; i < n; i++)
for(int j = 0; j < n; j++)
   arr[i][j] = (int*)calloc(sizeof(int));
}

[ssorower]

hi,
If you want to allocate for a 3D array of size n x n x n then every time you allocate you need to allocate for n number of intergers, that is the parameter for caloc function should be like, n*sizeof(int).

Your first allocation is OK, just chage the parameter as n*sizeof(int). But inside the first for loop you need to allocate one more time like,
arr[i]= (int**) calloc(n*sizeof(int));

And also, chage the parameter of the third allcation as n*sizeof(int). Thats it.

To summarize,
first you need to allocate for ***arr to hold n number of **arr[i],
then for each of them you need to allocate to hold n number of *arr[i][j]
and only then you can allocate for each of them to hold n number of a[i][j][k], that is integers.

Hope it helps :) and you probably should be able t write yourself now.

Thanks,
Sorower

[rsteph]

Sorry that this is a little off topic from your post, but I noticed in your code that you have instances with three *'s by the variable names, and type declarations. Can I ask why you use 3 *'s instead of just 1? I'm just getting back into programming, and into C++ programming, and that just struck me as something I hadn't seen before. Again sorry to get off topic for a moment.

[Silent1Mezzo]

Sorry that this is a little off topic from your post, but I noticed in your code that you have instances with three *'s by the variable names, and type declarations. Can I ask why you use 3 *'s instead of just 1? I'm just getting back into programming, and into C++ programming, and that just struck me as something I hadn't seen before. Again sorry to get off topic for a moment.

int* var; <---a pointer to an int
int*** var; <---a pointer to a pointer to a pointer of an int

Kind of like var[] vs var[][][]

[rsteph]

Ahh thank you. I'm still a little hazy on the advantages of that, but I can see what you're getting at as far as how it works. Thanks for the information.

[Silent1Mezzo]

Ahh thank you. I'm still a little hazy on the advantages of that, but I can see what you're getting at as far as how it works. Thanks for the information.

The main advantage is that with arrays you need to know the number of elements when you declare them but with pointers you can just allocate more memory and dynamically grow (or shrink) them.

[weaknessforcats]

Everyone is making this too hard. There are no 3D arrays in C or C++. All there are are one dimensional arrays. The first index specifies the number of elements:

int arr[3];       //has 3 elements. Each element an int
int arr[3][7];   //has 3 elements. Each element an array of 7 int
int arr[3][4][5];  //has 3 elements. Each element an array 4 arrays of 5 ints.
 

Further, arrays must be contiguous in memory. Element 0 then element 1, etc.

So you allocate int arr[3][4[5] as:

int* arr = malloc(3 * 4 * 5 * sizeof(int));
 

You now have 60 ints. If you want to use this as a 3D array then tou need to cast the return from malloc to the correct pointer. In this case, a pointer to a 4x5 array of int:

int (* arr)[4][5] =(int (*)[4][5] )malloc(3 * 4 * 5 * sizeof(int));
 

Remember, the name of an array is always the address of element 0. In this case the elements are 4x5 arrays. Using an int*** is a pointer to a pointer to a pointer to a single int. Trying to use this as a 3D array will produce the wrong address calculations.