How do I fix this convert error? Is this a bug?

insane

I have an integer and it's value is -16777216
I want to convert it to a float once done the float equals -1.677722E+07
Now if I convert the float back at this point everything works and I get a proper result.

However if I manually make the float equal -1.677722E+07 then convert back to int I get
a wrong result of -16777220 what is going on and how do I fix this? Example shown below in c# code.

int i = -16777216;
float f = Convert.ToSingle(i);                        // = -1.677722E+07 (works converting back)
label1.Text = i.ToString();
label2.Text = f.ToString();
f = -1.677722E+07f;                                   //Putting it in as a value makes error result
label3.Text = Convert.ToInt32(f).ToString();   //converting back

4 Replies

Know the answer? Post it — somebody with the same question will find it here.

Vulpes13y ago
Accepted answer
This sort of anomaly often occurs when converting between integral and floating point types and is not a bug as such.

Although they may appear to be the same (because they have the same string representation), the float you get when converting -16777216 to float is not the same as the float you get when you assign -1.677722E+07F to a float variable!

This can be seen by running the following code:

int i = -16777216;
float f = Convert.ToSingle(i);
float g = -1.677722E+07F;
Console.WriteLine( f == g);
foreach(byte b in BitConverter.GetBytes(f)) Console.Write("{0} ", b);
Console.WriteLine();
foreach(byte b in BitConverter.GetBytes(g)) Console.Write("{0} ", b);
Console.WriteLine();

The output is :

False
0 0 128 203
2 0 128 203

So, the two floats differ in the first byte.

Double is better than float in this respect as it has a precision of 15 or 16 digits compared to float's 7 digits.

However, if you're writing an application which can't tolerate this sort of anomaly it's best to use the decimal data type, rather than float or double, which is stored as a scaled integer accurate to 28 digits. The drawback is that it needs 16 bytes of storage (float requires 4 and double 8) and, when doing a lot of calculations, is much slower.
insane13y ago
I tried each of the codes above and it still gives -16777220. Are you getting a different result?
Veena Sarda13y ago

The +07f is the problem area. You are not representing the scientific notation correctly.
Try as shown in the screenshot
Thanks
Veena
Blocked Account13y ago
Hi,

Integer To Float Conversion:

int intNo = 2345;

float no = intNo;

Console.Write(no); // float can handle int because of upcasting

Float To Integer Conversion:

float floNo = 1.5f;

int num = (int)floNo; // but here .5 value will be lost due to downcasting

Console.Write(num);
so when you converting back float to int, it can be some value loss. this is the concept.
i have made some changes to your code.
int i = -16777216;
            float f = Convert.ToSingle(i);                        // = -1.677722E+07 (works converting back)
            Label1.Text = i.ToString();
            Label2.Text = f.ToString();
            //f = -1.677722E+07f;                                   //Putting it in as a value makes error result
            Label3.Text = ((int)(f)).ToString();   //converting back
try this