• jicman (8/8) Mar 09 2007 Greetings!
• Bill Baxter (13/24) Mar 09 2007 The lack of such a function probably has something to do with the fact

jicman <cabrera_ _wrc.xerox.com> writes:

Greetings!

Forgive me for this stupid question, but I have searched all over digitalmars
and std.math and I can not find a round(real,places) where real is a real
number and places is an integer suggesting the decimal places to round to.  I
had to do a subroutine for this using std.math.ceil() and some multiplication.
 But, is there a function in std.math, or anywhere to do this?

thanks,

josé

Bill Baxter <dnewsgroup billbaxter.com> writes:

jicman wrote:
 Greetings!
 
 Forgive me for this stupid question, but I have searched all over digitalmars
 and std.math and I can not find a round(real,places) where real is a real
 number and places is an integer suggesting the decimal places to round to.  I
 had to do a subroutine for this using std.math.ceil() and some multiplication.
  But, is there a function in std.math, or anywhere to do this?
 
 thanks,
 
 jos�

The lack of such a function probably has something to do with the fact 
that you can't represent 2.3 exactly in floating point.  But integers do 
have an exact representation.

There is a round() function in std.math though.  Something like this 
should work reasonably, as long as you don't let 'places' go too high.


static import std.math;
real round(real x, uint places)
{
    real pwr = pow(10.0,places);
    return std.math.round(x * pwr) / pwr;
}


--bb