Indexing matrix element member of struct - c++

Question

I'm trying to index an element from a matrix, which has been declared by creating a vector of vectors inside a struct.
Here's my code:
#include <vector>
using namespace std;
const int MAX_DIM = 1000;
struct TestStruct{
int MAX_DIM;
vector<vector<int> > matrix (int MAX_DIM, vector<int>(int MAX_DIM));
};
int main(){
TestStruct ts;
ts.MAX_DIM = 100;
ts.matriz[0][0] = 1;
return 0;
}
While compiling I've recieved the following error:
test.cpp:14:17: error: invalid types ‘<unresolved overloaded function type>[int]’ for array subscript
mierda.matriz[0][0] = 1;
Btw, do you know any other "cleaner" way to declare the matrix (without using anything else than vector from the class Vector)?
Thanks in advance!

Answer 1

The problem in your code and your logic is that you think the line:
vector<vector<int> > matrix (int MAX_DIM, vector<int>(int MAX_DIM));
Creates a matrix inside your struct.
In fact, you just declared a function named matrix that returns a vector<vector<int>. To have a member object inside your struct, you need to delete the parentheses, as such:
vector<vector<int> > matrix;
I assume you want to first specify the MAX_DIM value and then create a matrix based on that value. I would either recommend creating a constructor that takes an int as argument to assign that value to MAX_DIM and then create a matrix based on that value, or to create an initialise() function that needs to be called before you access the matrix.
First, preferable solution:
using namespace std;
const int MAX_DIM = 1000;
struct TestStruct{
TestStruct(int MD){
MAX_DIM = MD;
matrix = vector<vector<int>>(MD, vector<int>(MD));
}
int MAX_DIM;
vector<vector<int>> matrix;
};
int main(){
TestStruct ts(100); // matrix 100 by 100
ts.matrix[0][0] = 1; // works
return 0;
}
Or, if you don't want to deal with constructors:
using namespace std;
const int MAX_DIM = 1000;
struct TestStruct{
int MAX_DIM;
vector<vector<int>> matrix;
void initialise(){
matrix = vector<vector<int>>(MAX_DIM, vector<int>(MAX_DIM));
}
};
int main(){
TestStruct ts;
ts.MAX_DIM = 100; // specify the size
ts.initialise(); // construct the matrix
ts.matrix[0][0] = 1; // works again
return 0;
}
Note that the first solution requires less code and is more readable inside the main function. If you choose the second one, it will be easy to forget to either assign value to MAX_DIM or to call initialise().