KL Class Reference

Kullback-Leibler distance. More...

#include <distance.h>

---

Detailed Description

Kullback-Leibler distance.

Kullback-Leibler distance when talking about probability vectors (for interaction similarity)

class KL:public dist{ public: Constructor* KL(){type = 5;};

Compute the similarity

Parameters:

C1	First cluster to compare
C2	Second cluster to compare

Returns:

KL distance

template <typename t>=""> double operator() (T C1, T C2){ typename T::iterator it_t, it_t2;

double sum = 0;

for (it_t = C1.begin(), it_t2 = C2.begin(); it_t != C1.end() && it_t2 != C2.end(); it_t++, it_t2++){ sum += (*it_t) * (log((double)(*it_t)/(double)(*it_t2))); }

return sum; } };

---

The documentation for this class was generated from the following file:

• distance.h