Floating point is very weird in C language

Question

Floating point is weird in C language. I know it is generally used as (float)a/b. However, I am curious what is the main cause and reason for the code phenomenon below.

#include <stdio.h>

int main(void)
{
  int a=30, b=16;
  double divresult;
  divresult = a/b;  
  printf("result1 : %f \n", a/b);
  printf("result1 : %f \n", a/b);
  printf("result1 : %f \n", a/b);
  printf("result1 : %f \n", a/b);
  printf("result1 : %f \n", a/b);
  printf("result1 : %f \n", a/b);
  printf("result1 : %f \n", a/b);
  printf("result1 : %f \n", a/b);
  printf("result1 : %f \n", a/b);
  printf("divresult : %f \n",divresult);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  return 0;
}

output here

result1 : 0.000000 
result1 : 0.000000 
result1 : 0.000000 
result1 : 0.000000 
result1 : 0.000000 
result1 : 0.000000 
result1 : 0.000000 
result1 : 0.000000 
result1 : 0.000000 
divresult : 1.000000 
result2 : 1.000000 
result2 : 1.000000 
result2 : 1.000000 
result2 : 1.000000 
result2 : 1.000000 
result2 : 1.000000 
result2 : 1.000000 
result2 : 1.000000 
result2 : 1.000000 
result2 : 1.000000 
result2 : 1.000000 
result2 : 1.000000 
result2 : 1.000000

Why is this happening? I can't really understand. In general, we know that computers cannot accurately represent floating point numbers. But in this case, there seems to be some pattern. How does the change of values before and after "divresult printf" appear?

`a/b` results in an integer, but you try to print it using `%f` — perivesta, Feb 07 '21 at 13:22
%f is the format specifier for floats, and integer division results in an integer. — Mansoor, Feb 07 '21 at 13:24
@πάνταῥεῖ it is the wrong duplicate. The case of the problem is completely different — 0___________, Feb 07 '21 at 13:29

0___________ · Accepted Answer · 2021-02-07T13:35:11.107

Printf is supposed to print double but you pass integer. It is undefined behaviour. In this case it expresses itself by printing "remains" from the previous parameter passed. In this modified example it is clearly visible. But of course it is an UB and the behaviour of the program can be completely different. Also remember that the result of the integer division is also an integer. 30/16 == 1 in C and C++

(double)(a/b) is 1.0 as it converts the result of integer division to double

(double)a/b converts a to double and then does the double division.

#include <stdio.h>

int main(void)
{
  int a=30, b=16;
  double divresult;
  divresult = a/b;  
  printf("result1 : %f \n", (double)(a/b));
  printf("result1 : %f \n", a/b);
  printf("result1 : %f \n", a/b);
  printf("result1 : %f \n", (double)a/b);
  printf("result1 : %f \n", a/b);
  printf("result1 : %f \n", a/b);
  printf("result1 : %f \n", a/b);
  printf("result1 : %f \n", a/b);
  printf("result1 : %f \n", a/b);
  printf("divresult : %f \n",divresult);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", 0.0);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", (double)a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", (double)(a/b));
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", 0.0);
  printf("result2 : %f \n", a/b);
  printf("result2 : %f \n", a/b);
  return 0;
}

Result

x86-64 gcc 10.2

Program returned: 0
Program stdout

result1 : 1.000000 
result1 : 1.000000 
result1 : 1.000000 
result1 : 1.875000 
result1 : 1.875000 
result1 : 1.875000 
result1 : 1.875000 
result1 : 1.875000 
result1 : 1.875000 
divresult : 1.000000 
result2 : 1.000000 
result2 : 1.000000 
result2 : 0.000000 
result2 : 0.000000 
result2 : 0.000000 
result2 : 1.875000 
result2 : 1.875000 
result2 : 1.875000 
result2 : 1.875000 
result2 : 1.000000 
result2 : 1.000000 
result2 : 1.000000 
result2 : 0.000000 
result2 : 0.000000 
result2 : 0.000000

Floating point is very weird in C language

1 Answers1