Problem understanding something FLOAT

[mohibouki]

I am new to C++ so please don’t flame me :)

I wrote a simple prog. to test this course but something isn’t really working well and I can’t figure out why…

#include<iostream>
#include<string>
#include <iomanip> // for setprecision()
using namespace std;

main()
{
cout<<setprecision(7); //7 decimals
float v = 1;
float j = 3;
float cc;
cc = v/j;

//TEST with FLOAT NUMMERS
float ff = 0.3333333; // 7 decimals as set in "setprecusion(7);"

if(cc<ff) {
cout<<"cc is smaller then ff"<<endl;
}
else if(cc>ff) {
cout<<"cc is bigger then ff"<<endl;
}
else{
cout<<"cc equals to ff"<<endl;
}
cout<<cc<<" = cc"<<endl;
cout<<ff<<" = ff"<<endl;

return 0;
}

The output gives me that cc is bigger than ff…
I don’t understand why as I set precision to 7 and my var ff has also 7 decimals.
They should both be equal to eachother.

Any suggestions where I made an error?
Thanks!!

[Peter87]

setprecision changes how the number is printed. It does not affect the actual numbers and how they are compared.

[hacker804]

Its not actually performing float division
Try changing

float v = 1.0;
float j = 3.0;

Then check it.

[Peter87]

@hacker804
v and j are floats so it is performing float division.

[mohibouki]

I did both (originally it was

float v = 1.0;
float j = 3.0;

I tested it with 1 and 3 and had the same result as I defined them as floats.

Peter87 is absolutely correct: what you see isn't what you actually get...
I have set cout<<setprecision(30) and I get different numbers after the 7th digit.

I guess I need to be careful using == when comparing floats :)

[Peter87]

Yes, testing equality on floating point numbers is not to recommend. What you could do is to check if two numbers are almost equal.

if (std::abs(cc - ff) < a_very_small_number)
{
	// Almost equal.
}

If you want 7 decimals precision you can use 0.0000001 as the value of a_very_small_number.

[mohibouki]

Ha, nice work around.
Thanks Peter!