CODEWITHSUNDEEP
c++typesvector
Miek
Miek

Reputation: 1137

declare a vector of some unknown precision

I want to declare a vector in my .h file but depending on the precision of the data I send in I might want the vector to be of type double or I might want it to be of type float.

//tolerances.h

class verySimple{

public:
  verySimple();
  ~verySimple();

  void processTolerances(std::vector<double or float> tolerances);
};

Could I get a quick lesson, please? Thanks.

Upvotes: 0

Views: 129

Answers (3)

molbdnilo
molbdnilo

Reputation: 66449

It isn't entirely clear what you want to accomplish, but judging by your example you want a function that can accept either float or double vectors.

You can overload functions in C++:

class verySimple{
    public:
      void processTolerances(std::vector<float> tolerances);
      void processTolerances(std::vector<double> tolerances);
};

Then the appropriate one will be called depending on the argument.

Or, you can make the function a template:

template<typename T>
void processTolerances(std::vector<T> tolerances);

Or, if verySimple as a whole depends on the type:

template<typename T>
class verySimple{
    public:
      void processTolerances(std::vector<T> tolerances);
};

Which way to choose depends on the natures of verySimple and processTolerances - there's no general "best solution".

Upvotes: 0

DRVic
DRVic

Reputation: 2491

And if you're new to templates get everything working with something that isn't a template (either float or double) before converting it to a template. Most compiler warnings and errors are less informative when templates are involved.

Upvotes: 0

StilesCrisis
StilesCrisis

Reputation: 16300

This is what template is for.

You can say template <typename T> class verySimple and then void processTolorances(std::vector<T> tolorances); to do what you want.

You should check out a tutorial though as templates have a lot of gotchas for the uninitiated.

Upvotes: 4

Related Questions