Limiting decimal places of a float when converting to string in C++ - c++

Question

I'm making a function which prints the elements of a std::vector<float>.
Working code:
std::vector<float> components { 1, 2, 3 };
string result = "<";
for ( auto it = begin(this->components); it != end(this->components); ++it ) {
result.append(to_string(*it));
if (it != (this->components))
result.append(", ");
}
result.append(">");
std::cout << result;
The intended result is if "components" has elements 1,2,3, for example, it will print: <1, 2, 3>.
Right now it is printing the numbers as floats, of course, like < 1.000000, 2.000000, 3.000000, >.
Is there a way I can control how many decimal places are put into the string, without having to manually go through it character by character?
As a side note, how do I prevent it from adding a ',' after the last element?

Answer 1

You may use std::stringstream.precision for that.
Just create a std::stringstream convert it to a string and your done.
Like so:
stringstream ss;
ss.precision(3);
ss << "<";
for ( auto it = begin(this->components); it != end(this->components); ++it ) {
ss << *it;
if (it != (this->components))
ss << ", ";
}
ss << ">";
string result = ss.str();

Answer 2

you can use sprintf() before casting:
float a = 1.000000;
char aa[20];
sprintf(aa, "%1.3f", a);
Here is the complete code which i ran:
#include <vector>
#include <iterator>
#include <iostream>
using namespace std;
int main()
{
std::vector<float> components{ 1, 2, 3 };
string result = "<";
for (auto it = components.begin(); it != components.end(); ++it) {
float a = *it;
char aa[20];
sprintf(aa, "%1.3f", a);
result.append(string(aa));
if (it+1 != components.end())
result.append(", ");
}
result.append(">");
std::cout << result.c_str();
getchar();
return 0;
}
Output:

Answer 3

I would do it like this using a stringstream.
#include <iostream>
#include <sstream>
#include <iomanip>
#include <vector>
int main()
{
std::vector<float> components {1, 2, 3, 1.5f, 2.5f, 3.5f, 1.25f, 2.25f, 3.25f, 1.12345f};
std::stringstream result;
result << "<";
for(auto it = std::begin(components); it != std::end(components); ++it)
{
if(it != std::begin(components))
{
result << ", ";
}
result << *it;
}
result << ">";
std::cout << result.str();
return 0;
}
You could also use std::fixed and std::setprecision to modify the output further as you desire.
Printing the comma before the next item for all but the first item fixes your trailing comma issue.
Here's a demo of it working:

Answer 4

As #Axalo already noted, you can use setprecision with an ostream to set its precision (and it can be used with any ostream, not just cout).
To eliminate the trailing comma, I'd probably use the infix iterator I've posted elsewhere.
Using that, the code could be written something like this:
#include <iostream>
#include <sstream>
#include <vector>
#include <iomanip>
#include "infix_iterator.h"
int main () {
// Data that would display extra precision if we didn't stop it
std::vector<float> components { 1.123f, 2.234f, 3.345f };
std::ostringstream buff("<", std::ios::app);
buff << std::setprecision(2);
std::copy(components.begin(), components.end(),
infix_ostream_iterator<float>(buff, ", "));
buff << ">";
std::cout << buff.str();
}
Result: <1.1, 2.2, 3.3>