Simple Rounding Function for ansi C (using double)

[Darren Schehl]

Question: http://bytes.com/topic/c/answers/217...nding-function

Is there a std lib rounding function that will round a real to a
given number of decimal places? Something like:

double round_it(double number, int decimal_digits)

pass 34.5678, 3 to it and it returns 34.568

Note: posting solution to archived question:

[donbock]

You didn't ask a question. Does this function do what you want it to do?

[Darren Schehl]

double d_round(double in_d,int r2dec)
{ double d_whl=0,d_dec=0,d_last=0,d_rnd=0,d_rlst=0;
  int x, to_dec =10;
for (x = 0; x<r2dec; x++)
    {to_dec = to_dec*10;}
d_whl = floor(in_d);
d_dec = in_d - d_whl;
d_dec = floor(d_dec*to_dec);
d_last = floor(d_dec/10)*10;
d_rnd = d_dec-d_last;
if(d_rnd>=5)
  d_rlst = d_whl + ((d_last+10)/to_dec);
else
  d_rlst = d_whl + (d_last/to_dec);
return d_rlst;
}

note: respond to: http://bytes.com/topic/c/answers/217...nding-function

[Darren Schehl]

This one even works better:
Rounds whole numbers with neg numbers (left of dec)
Rounds dec with positive numbers (right of dec)

#include <math.h>
double round2(double number, int r2dec)
{
 double d_rtn, i_div, i_mlt;
    i_div = pow(10, r2dec+1);
    i_mlt = pow(10, r2dec);
    d_rtn = floor((number + (5/i_div))*(i_mlt))/(i_mlt);
 return d_rtn;
}

[Darren Schehl]

This one handles negative numbers also.
#include <math.h>
double round_dbl(double number, int r2dec)
{double d_rtn, i_div, i_mlt;
int half = 5;
if(number<0) half = -5;
i_div = half/pow(10,r2dec+1);
i_mlt = pow(10, r2dec);
d_rtn = int((number + i_div)*(i_mlt))/(i_mlt);
return d_rtn;}

[donbock]

The following standard library functions are available in C99 to help you with rounding:

ceil() Rounds towards +infinity.
floor() Rounds towards -infinity.
round() Rounds to nearest integer.
trunc() Rounds towards zero.
nearbyint() Rounds as per the current rounding direction.
rint() same as nearbyint(), but handles errors differently
fesetround() Set the current rounding direction.

All of these functions round to the nearest integer. It is a simple matter to round to any decimal digit:

double y = round(x*100)/100;

However, inexact floating-point representation may get in your way.

[Darren Schehl]

round() doesn't exist in MS Visual C++ 6.0