CODEWITHSUNDEEP
javacasting
Vitaly Zdanevich
Vitaly Zdanevich

Reputation: 14812

Why do I need to cast a Float to float before cast to byte?

Float f = new Float(10.71);    
byte b = (byte)f;

I get

Cannot cast from Float to byte

Why do I have to do this?

byte b = (byte)(float)f;

Upvotes: 4

Views: 4709

Answers (5)

Alex
Alex

Reputation: 1008

Be aware on the differences between classes and primitives.

You'll never be allowed to cast any object to a primitive. These are different entities as an object creation pass through an instantiation, hence JVM treat objects and primitives as different entities without relation between them.

The classes related to primitives (Integer, Float, ...) are used to wrap these primitives and provide some manipulation methods. Furthermore, this classes have a specific cast procedure that allows then to return the primitive they are wrapping.

Only for these objects

float myFloat = (float)f;

is equivalent to

float myFloat = f.floatValue();

Once you have your object transformed to a float type, you can cast it backwards to an int (which will remove the decimal digits obviously).

int myInt = (int)(float)f;
int mySameInt = (int)f.floatValue();     //same value than above

Upvotes: 1

Adam Arold
Adam Arold

Reputation: 30528

You cannot cast an Object to a primitive. You can only cast Float to float because of java's autoboxing feature.

This "feature" is being assaulted endlessly by angry developers but I guess it is here to stay for backwards compatibility reasons.

I don't like it either but you have to learn to live with it. Just don't use Longs or Floats if you can because it is not funny when you try to do computation with null values.

Upvotes: 11

vogomatix
vogomatix

Reputation: 5041

Not entirely certain but can't you do

byte bo = f1.byteValue();

Upvotes: 0

Silviu Burcea
Silviu Burcea

Reputation: 5348

You cannot cast an Object to a primitive.

However, you can use the byteValue() method from Number class.

Upvotes: 0

Sergey Kalinichenko
Sergey Kalinichenko

Reputation: 726539

Since Float is a "boxed" float, you cannot cast it directly. However, in your situation you do not need casting at all - you can go directly to byte, like this:

byte b0 = f1.byteValue();

The above is possible because Float is a subclass of java.lang.Number, which provides the byteValue() method.

Upvotes: 4

Related Questions