Thread: Array size at run time unknown

Hello All,

I'm looking to write a program which will read in a set of data points from a file and place these in a float-array, the problem is i don't know how many data points i will have. I have been making oversized arrays and then just filling them up to N values and keeping track of N but I would like a more dynamic way of doing this. I have looked at the vector class but I am using a third-party library in my program and i need to pass float arrays to some of it's functions otherwise the vector would be the perfect solution.

is there a way to initailise an array dynamically or initalise an array with a vector ?
How does the vector class get around the issue of dynamic allocation?

I'd say a vector is definitely the way to go. You can use a vector and still pass the array to the third party library:

Code:

std::vector<float> data(5, 3.14);
thirdPartyFunction(&data[0], data.size());

The &data[0] part provides a pointer that will work just like a pointer to a normal array.

Magic. It works !

Thanks you both.

Just make sure the size of the vector is big enough for what the function expects. For example, if the function is expecting an array with three elements, you can't use that trick with an empty vector:

Code:

std::vector<float> data;
fillArrayWithThreeValues(&data[0]); // Bad!