C++ :: 2 Dimensional Vector Represented As A Class

I have to create a class to represent a 2 dimensional vector. I need to include certain member functions such as a function to find magnitude of the vector, and one to find the dot product of that vector with another vector, and several others too. That's all fine. A stipulation of the problem is that I must include a constructor which can take cartesian form of the vector and a constructor which can take polar form of vector. Since this involves overloading the constructor the best solution I have come up with is to create the object with either doubles or floats so that the compiler can choose the correct constructor. This seems like a really bad idea. Is there a way I can get the compiler to choose the correct constructor without doing it using the precision? Here is a sample of my header file, there are many more member functions

class Vector{
public:
Vector(double x, double y, double v, double w){
myArray[0]=x;
myArray[1]=y;
additionalArray[0]=v;
additionalArray[1]=w;

[Code] .....

In my .cpp file the object is created with either four doubles or four floats depending on which constructor I want to implement. There must be a better way. Additional Array is created for use in member function which require calculations with a second 2d vector.