Floating point arithmetic operations

Let's perform floating-point arithmetic operations . Adds, subtracts, multiplies, and divides floating point numbers.

64-bit floating-point arithmetic operations using double type

It is a 64-bit floating point four arithmetic operation using double type. If you want to output floating point, use "%f" for the format specifier.

#include <stdio.h>

int main (void) {
  double num1 = 1.45;
  double num2 = 4.67;
  
  // Addition
  double num_add = num1 + num2;
  
  // Subtraction
  double num_sub = num1 --num2;
  
  // Multiplication
  double num_mul = num1 * num2;
  
  // Division
  double num_div = num1 / num2;
  
  printf("Add%f\n", num_add);
  printf("Subtract%f\n", num_sub);
  printf("Multiply%f\n", num_mul);
  printf("Division%f\n", num_div);
}

Output result.

Add 6.120000
Subtract -3.220000
Multiply 6.771500
Division 0.310493

32-bit floating point four arithmetic operations using float type

It is a 32-bit floating point four arithmetic operation using float type. I'm using the floating point literal f suffix to make it a float type. If you want to output floating point, use "%f" for the format specifier.

#include <stdio.h>

int main (void) {
  float num1 = 1.45f;
  float num2 = 4.67f;
  
  // Addition
  float num_add = num1 + num2;
  
  // Subtraction
  float num_sub = num1 --num2;
  
  // Multiplication
  float num_mul = num1 * num2;
  
  // Division
  float num_div = num1 / num2;
  
  printf("Add%f\n", num_add);
  printf("Subtract%f\n", num_sub);
  printf("Multiply%f\n", num_mul);
  printf("Division%f\n", num_div);
}

Output result.

Add 6.120000
Subtract -3.220000
Multiply 6.771500
Division 0.310493

Features of floating point arithmetic

It is a feature of floating point arithmetic.

There is an error, not an exact value

Floating point has an internal error, except for numbers that can be represented in binary.

For example, 0.5, 0.25, 0.125, etc. can be expressed as 0.1, 0.01, 0.001 in binary, so there is no error, but if it is not, there is an error from the exact value handled in mathematics. increase.

Floating-point arithmetic does not result in an error even if it is divided by zero

Floating-point arithmetic, unlike integer arithmetic, does not result in an error when divided by zero. A value called nan (Not A Number) is set. There is no need to check before calculation like integer division.

Can express the value of infinity

If the maximum value that can be expressed by double or float is exceeded, an infinite value called Inf is set. Depending on the sign, there are positive infinity and negative infinity.

Even if the same 0 exists, there are 0s with different signs

Even with the same 0, there are 0s with a positive sign and 0s with a negative sign. Positive 0s and negative 0s are considered equivalent in comparison operations, but they have different internal representations and different results in conscious operations.

Calculation results may differ depending on the CPU

The double type represents a 64-bit floating point number, but the CPU may actually calculate beyond the 64-bit precision. As an example, Intel CPU has an area for 80bit floating point calculation.

In other words, it may differ from the originally expected calculation result as a 64-bit floating point number.

C Language Zemi