C library - fesetround() function

---

---

The C fenv library fesetround() function sets the rounding mode for arithmetic operations(floating-point). While performing this function to real world application it require specific calculation based on floating-point such as banking/financial software always round the value towards zero(interest calculation).

Syntax

Following is the C library syntax of the fesetround() function.

int fesetround(int round);

Parameters

This function accepts only a single parameter which is an interger. Below is the list of following parameter that determine the desired value(round) −

• FE_TONEAREST: This parameter define the round to nearest which is zero.

• FE_DOWNWARD: This parameter define the round value toward negative infinity.

• FE_UPWARD: This parameter define the round value toward positive infinity.

• FE_TOWARDZERO: This parameter define the round value zero.

Return Value

• This function returns an integer value(zero), if it is successful. If the parameter is not a supported rounding mode, a nonzero value is returned.

Example 1

Following is basic C library program set the parameter FE_TONEAREST to the function fesetround() to see its demonstration.

#include <stdio.h>
#include <fenv.h>
#include <math.h>
int main() {
   double value = 9.5;
   double results[4];
   
   // Set rounding modes and perform calculations
   fesetround(FE_TONEAREST);
   results[0] = nearbyint(value);
   
   fesetround(FE_DOWNWARD);
   results[1] = nearbyint(value);
   
   fesetround(FE_UPWARD);
   results[2] = nearbyint(value);
   
   fesetround(FE_TOWARDZERO);
   results[3] = nearbyint(value);
   
   // Print results
   printf("Rounding FE_TONEAREST: %f\n", results[0]);
   printf("Rounding FE_DOWNWARD: %f\n", results[1]);
   printf("Rounding FE_UPWARD: %f\n", results[2]);
   printf("Rounding FE_TOWARDZERO: %f\n", results[3]);
   
   return 0;
}

Output

The above code produces the following result −

Rounding mode set to FE_TONEAREST successfully.

Example 2

Below the program demonstrates each and every parameter that compares the value of rounded mode with a single calculation.

#include <stdio.h>
#include <fenv.h>
#include <math.h>
int main() {
   double value = 9.5;
   double results[4];
   
   // Set rounding modes and perform calculations
   fesetround(FE_TONEAREST);
   results[0] = nearbyint(value);
   
   fesetround(FE_DOWNWARD);
   results[1] = nearbyint(value);
   
   fesetround(FE_UPWARD);
   results[2] = nearbyint(value);
   
   fesetround(FE_TOWARDZERO);
   results[3] = nearbyint(value);
   
   // Print results
   printf("Rounding FE_TONEAREST: %f\n", results[0]);
   printf("Rounding FE_DOWNWARD: %f\n", results[1]);
   printf("Rounding FE_UPWARD: %f\n", results[2]);
   printf("Rounding FE_TOWARDZERO: %f\n", results[3]);
   
   return 0;
}

Output

On execution of above code, we get the following result −

Rounding FE_TONEAREST: 10.000000
Rounding FE_DOWNWARD: 9.000000
Rounding FE_UPWARD: 10.000000
Rounding FE_TOWARDZERO: 9.000000

Example 3

In this program, we set the round mode(floating-point) to perform the task of division.

#include <stdio.h>
#include <fenv.h>

int main() {
   // Set rounding mode to upward
   fesetround(FE_UPWARD);
   
   // Perform a division
   double result = 10.0 / 3.0;
   printf("Result with FE_UPWARD: %f\n", result);
   
   return 0;
}

Output

After executing the above code, we get the following result −

Result with FE_UPWARD: 3.333334