问题
When performing conversion from byte to double by mathematical operations I've got expected result.
Performing the same mathematical operations to reverse the changes in value, results in correct value as long as I keep double type.
But when at the end I converts double result back to byte value the conversion results is incorrect by 1.
This is the case only for some byte values.
Exact process:
byte b = 82; Console.WriteLine(b); // initial byte value
double d = (b / 100.0) + 2.00; Console.WriteLine(d); // 82 / 100 -> 0.82 + 2.00 -> 2.82 OK
double dt = d - 2.00; Console.WriteLine(dt); // 2.82 - 2.0 -> 0.82 OK
double db = dt * 100.0; Console.WriteLine(db); // 0.82 * 100 -> 82 (double) OK
byte dbb = (byte)db; Console.WriteLine(dbb); // (byte)82 -> 81 ERROR ERROR ERROR
b = Byte.Parse(db.ToString()); Console.WriteLine(b); // 82 -> "82" and to byte OK
Why it is happening?
The double result does not have any values after comma.
回答1:
if you debug your code you will see what is going on:
double d = (b / 100.0) + 2.00; Console.WriteLine(d)// => d = 2.82
double dt = d - 2.00; Console.WriteLine(dt) // => dt = 0.81999999999999984
double db = dt * 100.0; Console.WriteLine(db) // => db = 81.999999999999986
byte dbb = (byte)db; Console.WriteLine(dbb) //=> dbb = 81, because Byte is cut off after the ","
If you use decimal instead of double it will work out.
See: https://stackoverflow.com/questions/2741903/c-sharp-4-double-minus-double-giving-precision-problems
To make it complete:
decimal d = (b / 100.0m) + 2.00m; Console.WriteLine(d);
decimal dt = d - 2.00m; Console.WriteLine(dt);
decimal db = dt * (decimal)100.0; Console.WriteLine(db);
byte dbb = (byte)db; Console.WriteLine(dbb);
You can either cast (decimal) or use the "m" behind the value.
回答2:
Instead of byte dbb = (byte)db; use explicit rounding with byte dbb = (byte)Math.Round(db);.
来源：https://stackoverflow.com/questions/40673698/double-to-byte-type-conversion-error