CODEWITHSUNDEEP
c++doubleroundingostream
Joe Haring
Joe Haring

Reputation: 1

rounded the value of double

after searching an answer to my problem I decide to ask.

I was told I need to round the value of double into 3 decimal places.

I need to use this :

os.setf(std::ios::fixed,std::ios::floatfield);
os.precision(3);

so I am understand I need to create an object of type ostream so I wrote this:

ostream os;
os.setf(std::ios::fixed,std::ios::floatfield);
os.precision(3);
double a = 3.12364;
os << a;

but there is a compilation error so I changed the first row into :

ostream os(NULL);

then the compiler shut up, but I know it did not do anything. so how do I round the value using ostream ?

pay attention I do not want to print. I do not need/want to use cout

Upvotes: 0

Views: 640

Answers (3)

Potatoswatter
Potatoswatter

Reputation: 137810

Assuming you want to put the double into a std::string, use std::ostringstream (which is derived from std::ostream).

ostringstream os;
os.setf(std::ios::fixed,std::ios::floatfield);
os.precision(3);
double a = 3.12364;
os << a;

std::string rounded( os.str() ); // get the result

If you want to round it mathematically, then do

round( a * 1000 ) / 1000

Using a string in the middle of an equation is always a red flag ;v) .

Upvotes: 1

James Kanze
James Kanze

Reputation: 153919

First: if you need to use ostream, you don't want to round the value of a double to three decimal places, you want to display the value with three decimal places. (To round it, you'd use round( value * 100 ) / 100.) Second, you usually don't use an ostream object, you use an object of a type derived from ostream: usually an ofstream or ostringstream.

The question is where you want this rounded value to appear. If you need it in a file, then you will use ofstream; if you want it in a string (e.g. to place in a window component), it's ostringstream, and if you just want it to appear on standard out, just output it to cout (which is of a type derived from ostream).

Also, don't forget to restore the flags and the precision when you're done: setf and precision make permanent changes to the underlying stream.

Upvotes: 0

Masked Man
Masked Man

Reputation: 11025

If you do not really want to save the result in a string, you could use plain mathematics.

#include <cmath>

float myRound(float input) {
    return input > 0
        ? floor(input * 1000 + 0.5) / 1000
        : ceil(input * 1000 + 0.5) / 1000;
}

Be aware though, that the result will be incorrect if the multiplication by 1000 causes an overflow. In C++11, you can also make use of std::round.

Upvotes: 1

Related Questions