A more efficient loop structure?

[wmichaelv]

Hi, folks. I just started learning the C programming. I just finished a program to switch the 100 digit and 1 digit of a float number with the least amount of the two.

This is my program:

#include <stdio.h>
 
/**
 * Function Prototypes
 */
 
int modifyIntegralPartMichaelW();
 
int main() {
 
  double z;
 
  printf("\nTest #1:\n\n"
         "  Calling modifyIntegralPartMichaelW() --\n");
 
  modifyIntegralPartMichaelW();
 
  printf("\nTest #2:\n\n"
         "  Calling modifyIntegralPartMichaelW() --\n");
 
  modifyIntegralPartMichaelW();
 
  return 0;
}
 
/**
 * Function Definition
 */
 
int modifyIntegralPartMichaelW() {
 
  int v;
  int w;
  int x;
  int y;
  double z;
 
  printf("    Enter a floating-point: ");
  scanf("%lf", &z);
 
  w = z;
  x = w %10;
  v = x;
  y = w /100 %10;
  w = y;
  x = x * x;
  y = y * y;
 
  if (x > y) {
    z = z - v + w;
  } else {
      z = z - w * 100 + v *100;
  }
 
  printf("\n  Printing in main() : modified value : %.4lf\n\n", z);
 
  return (z);
}

Been browsing through the net for couple days, and stuck at "nested if" loop. The program is done, and proved to be working well. Just need to polish the appearance a bit, especially that loop. Any guidance is much appreciated.

[weaknessforcats]

You can't do this with a float.

You had to assign the float to an int, which you can't do since int has no decimal portion. This gives you an int that will not convert back to the original float.

The first thing to do is stick with int only.

However, now it is binary and not decimal. That is, there is no 100 digit to switch with a 1 digit.

[wmichaelv]

Thank you for the input. I'll start modifying the code right away.

[wmichaelv]

Thanks, weaknessforcats(?).

I realized that I couldn't take the 100 digit and 1 digit when the input is in float, thus I convert the input into int using a function.

In that function, the program then take the 100 digit and 1 digit of the intInput; compare them, and do some calculation with the float input so that the float output shows the desired result, which is changing the 100 digit and 1 digit with the smaller of the two.

I have finished modifying the code, changing the variables with better nicknames, and having some fun with the return function. Is there a better structure for this code?

#include <stdio.h>
 
/**
 * Function Prototypes
 */
 
double modifyIntegralPartMichaelW();
 
int main() {
 
  double doubleInput;
  double result;
 
  printf("\nTest #1:\n\n"
         "  Calling modifyIntegralPartMichaelW() --\n");
  printf("    Enter a floating-point: ");
  scanf("%lf", &doubleInput);
 
  result = modifyIntegralPartMichaelW(doubleInput);
 
  printf("\n  Printing in main() : modified value : %.4lf\n\n", result);
 
  printf("\nTest #2:\n\n"
         "  Calling modifyIntegralPartMichaelW() --\n");
  printf("    Enter a floating-point: ");
  scanf("%lf", &doubleInput);
 
  result = modifyIntegralPartMichaelW(doubleInput);
 
  printf("\n  Printing in main() : modified value : %.4lf\n\n", result);
 
  return 0;
}
 
/**
 * Function Definition
 */
 
double modifyIntegralPartMichaelW( double functionInput) {
 
  int intInput;
  int digit1;
  int digit100;
  double functionOutput;
 
  intInput = functionInput;
  digit1 = intInput % 10;
  digit100 = intInput / 100 % 10;
 
  if (digit1 * digit1 > digit100 * digit100) {
      functionOutput = functionInput - digit1 + digit100;
  } else {
      functionOutput = functionInput - digit100 * 100 + digit1 *100;
  }
 
  return functionOutput;
}

[divideby0]

Reading numbers of any type with scanf is risky. You might want to check its return value to make sure your number is valid before you enter your algorithm

if(scanf("%lf", &doubleInput) == 1)
{
   // proceed with program
}
else
{
   // deal with possible stream error
}

You can bypass the loop altogether by reading user input into a char array - you can keep the whole and fractional parts too.

char user_input[] = "876.543210";
int whole, fraction;
double output;
 
if((sscanf(user_input, "%d.%d", &whole, &fraction)) == 2)
{
    // algorithm to swap whole part
 
    // build the output
 
    sprintf(user_input, "%d.%d", whole, fraction);
    if((sscanf(user_input, "%lf", &output)) == 1)
        printf("number = %f", output);
}

[weaknessforcats]

Using a float like it was an array is not going to work. It's not an array and it does rounding so your answers may vary.

If you really need to switch the 100 digit with the 1 digit, then I suggest you convert your float input to a C string, which is a char array. You should be able to use ftoa to convert your double to a string.

Then you look for the decimal point. From there you can find the char in the 1 and 100 position. Make the swap. Then use atof to convert your string back to a float.

[wmichaelv]

Thanks for the great feedbacks!!

I'm in no hurry to modify this, so my update might come a bit too slow, but I'll be sure to experiment with every given method before posting the next code structure.