how to get all decimal number from string in c++

Question

I'm trying to get value (all decimal number) from a text in c++. But I have a problem and I couldn't solve it

#include "pch.h"
#include <iostream>
#include <fstream>
#include <sstream>
#include <string>

int main()
{
    std::ifstream infile("C:\\thefile.txt");
    float a, b;
    while (infile >> a >> b)
    {
        // process pair (a,b)
    }

    std::cout << a << " " << b;

}

thefile.txt:

34.123456789 77.987654321

When I run the above code,

a = 34.1235 
b = 77.9877

but I want

a = 34.123456789 
b = 77.987654321

what should I do?

EDIT: I don't want to print out of a and b. I just want they get the exact values.net

Does your computer have enough precision in a `float` to do that? Are you sure? — Lightness Races in Orbit, Jul 31 '19 at 13:27
@LightnessRacesinOrbit I am not sure anything. I don't want to print a and b. I just want a and b get the all values. If they already get all values and if it appear in a short way because of printing, it's okay. can you tell me ? — baybaybay, Jul 31 '19 at 13:31
@NathanOliver also which type should I use? double or float? — baybaybay, Jul 31 '19 at 13:33
You can check the real value of the variable using a *debugger*. But for printing, the duplicate answers it well. — Yksisarvinen, Jul 31 '19 at 13:33
@Yksisarvinen thank you. also float a,b; is correct? or should I use double? — baybaybay, Jul 31 '19 at 13:35
@NathanOliver Really? I thought a IEEE 754 single-precision float only guaranteed six-ish decimal significant figures (barring looser constraints on powers of two) — Lightness Races in Orbit, Jul 31 '19 at 13:39
@LightnessRacesinOrbit You're correct. Not sure what I was thinking. They'd need a double for full precision. — NathanOliver, Jul 31 '19 at 13:40
@baybaybay Your comment (written three times!) does not add anything to the discussion. — Lightness Races in Orbit, Jul 31 '19 at 15:31

score 5 · Answer 1 · answered Jul 31 '19 at 13:42

You can't.

A float can only give you six (ish) decimal significant figures. The values in your file, after conversion from string, cannot be held in a float.

First, you need to switch to double, otherwise you won't even have a variable with the full numerical value.

Then, for output, be careful to specify the precision you want.

Please remain aware of the foibles of floating-point, and consider sticking with strings, depending on what you're doing with this data.

score 1 · Answer 2 · answered Jul 31 '19 at 13:56

1

There are two main things i see in your code.

float precision is not enough for you data you need double data type.
you need to set proper cout precision to get your desired output.

This code will work for you.

#include <iostream>
#include <fstream>
#include <sstream>
#include <string>

int main()
{
    std::ifstream infile("newfile.txt");
    double a, b;
    std::cout.precision(11);
    while (infile >> a >> b)
    {
        // process pair (a,b)
    }

    std::cout << a << " " << b;
}

answered Jul 31 '19 at 13:56

GIRISH kuniyal

740
1
5
14

this doesn't show all value when I print out. But I don't want to show, I just want they get the exact values. Do they get the exact values if I use double? – baybaybay Jul 31 '19 at 14:25
@baybaybay Try it and find out? – Lightness Races in Orbit Jul 31 '19 at 15:32

score 0 · Answer 3 · answered Jul 31 '19 at 13:58

0

you should read your inputs into a string and then convert them into double (depending on enough precision in float on your computer). you can directly display strings. A possible answer is here.

string word;  
openfile >> word;
double lol = atof(word.c_str());

answered Jul 31 '19 at 13:58

Shafiqa Iqbal

158
5

Why do this yourself when IOStreams will do it for you? In code that the OP's written? – Lightness Races in Orbit Jul 31 '19 at 14:05
@ShafiqaIqbal I didn't get all what you said – baybaybay Jul 31 '19 at 14:29

how to get all decimal number from string in c++

3 Answers3