Allocating 2-D Arrays on Heap

Introduction

We’ve seen that allocating one-dimensional arrays is pretty straightforward. To allocate one array of 15 doubles, we can write the following:

double arr = malloc(15 * sizeof(double));

Then we can use this array just like we use any other array:

arr[0] = 5.0; //to write to the array
double x = arr[1]; //to read from the array

The question is, how can we dynamically allocate arrays with higher dimensions? We’ll present a strategy called “arrays of pointers to arrays” for the two-dimensional case. This strategy can extend to three-dimensional arrays and beyond.

Matrix structure

To understand this technique, let’s inspect how a 3x4 matrix looks:

As per the image above, we can view the matrix as multiple arrays stacked on top of each other, and each array is one row. Therefore, a matrix or two-dimensional array is an array of arrays.

To dynamically allocate a matrix with n lines and m columns, imagine dynamically allocating n individual one-dimensional arrays, where each array will serve as a row. We know how to allocate one-dimensional arrays, so this is doable.

For example, if we assume a 3x4 matrix that stores integers, we can do:

malloc(4 * sizeof(int)) //allocate first row array
malloc(4 * sizeof(int)) //allocate second row array
malloc(4 * sizeof(int) //allocate third row array
malloc(4 *

...