c++ - Result Difference between int and float in Generic Function -

compile , run code; compare result if using integers or floats fractional values. why there difference?

here code:

#include <iostream> using namespace std;  template<class t>  t find(t array[], t len, t num){         (int = 0; < len; ++i){         if (array[i] == num)               return i;     }     return -1; }  int main () {       int array1[5] = { 4, 7, 3, 5, 6 }, num1;     float array2[5] ={121.2, 111.5, 300.1, 500.1, 600.1 }, num2;      cout << "enter int:" << " " ;     cin >> num1;      cout << "enter float:" << " " ;     cin >> num2;              int x = find<int>(array1,5,num1);     float y= find<float>(array2,5,num2);       cout << "the index int is:" << "  " << x << endl;              cout << "the index float is:" << "  " << y << endl;      return 0; } 

i couldn't find difference between 2 results when using ints , when using floats.

the issue must not compare floats == due internal representation of floating point numbers. if using result expression, can mathematically correct find number in array, due limitations in floating point representation == not find it.

consider following code:

template<class t> int find(t array[], size_t len, t num){         (size_t = 0; < len; ++i){         cerr << std::setprecision(7) << "compare: " << array[i] << ", " << num;         if (array[i] == num) {              cerr << " => equal" << endl;              return i;         } else {              cerr << " =>not equal" << endl;         }     }     return -1; }   int main () {       float array2[2] ={500.1, 1.0 }, num2;      float = 500.1 / 2;     int = find<float>(array2, 2, 2*a);       = 1.0 / 0.3333;     = find<float>(array2, 2,  * 0.3333);       return 0; } 

output:

compare: 500.1, 500.1 => equal compare: 500.1, 0.9999999 =>not equal compare: 1, 0.9999999 =>not equal 

so, though 1.0 / 0.3333 * 0.3333 1.0 mathematically, not when using floating point arithmetics - 0.99999 instead. therefore == not treat numbers equal in last comparison.

to solve this, compare difference of 2 numbers small epsilon value. float.h explicitly defines constant flt_epsilon this:

if (fabs(array[i] - num) < flt_epsilon) { ...