What is the best practice when coding math class/functions ?
- by Isaac Clarke
Introductory note : I voluntarily chose a wide subject. You know that quote about learning a cat to fish, that's it. I don't need an answer to my question, I need an explanation and advice. I know you guys are good at this ;)
Hi guys,
I'm currently implementing some algorithms into an existing program. Long story short, I created a new class, "Adder". An Adder is a member of another class representing the physical object actually doing the calculus , which calls adder.calc() with its parameters (merely a list of objects to do the maths on).
To do these maths, I need some parameters, which do not exist outside of the class (but can be set, see below). They're neither config parameters nor members of other classes. These parameters are D1 and D2, distances, and three arrays of fixed size : alpha, beta, delta.
I know some of you are more comfortable reading code than reading text so here you go :
class Adder
{
public:
Adder();
virtual Adder::~Adder();
void set( float d1, float d2 );
void set( float d1, float d2, int alpha[N_MAX], int beta[N_MAX], int delta[N_MAX] );
// Snipped prototypes
float calc( List& ... );
// ...
inline float get_d1() { return d1_ ;};
inline float get_d2() { return d2_ ;};
private:
float d1_;
float d2_;
int alpha_[N_MAX]; // A #define N_MAX is done elsewhere
int beta_[N_MAX];
int delta_[N_MAX];
};
Since this object is used as a member of another class, it is declared in a *.h :
private:
Adder adder_;
By doing that, I couldn't initialize the arrays (alpha/beta/delta) directly in the constructor ( int T[3] = { 1, 2, 3 }; ), without having to iterate throughout the three arrays. I thought of putting them in static const, but I don't think that's the proper way of solving such problems.
My second guess was to use the constructor to initialize the arrays
Adder::Adder()
{
int alpha[N_MAX] = { 0, -60, -120, 180, 120, 60 };
int beta[N_MAX] = { 0, 0, 0, 0, 0, 0 };
int delta[N_MAX] = { 0, 0, 180, 180, 180, 0 };
set( 2.5, 0, alpha, beta, delta );
}
void Adder::set( float d1, float d2 ) {
if (d1 > 0)
d1_ = d1;
if (d2 > 0)
d2_ = d2;
}
void Adder::set( float d1, float d2, int alpha[N_MAX], int beta[N_MAX], int delta[N_MAX] ) {
set( d1, d2 );
for (int i = 0; i < N_MAX; ++i) {
alpha_[i] = alpha[i];
beta_[i] = beta[i];
delta_[i] = delta[i];
}
}
My question is : Would it be better to use another function - init() - which would initialize arrays ? Or is there a better way of doing that ?
My bonus question is : Did you see some mistakes or bad practice along the way ?
