Numeric Conversions

You can convert between different numeric types. Some of these conversions must be explicitly stated in your code, and others can be performed implicitly. In general, conversions from a smaller type to a larger type (widening conversions) can happen automatically because they shouldn't cause a loss of information. Conversions from a larger type to a smaller type will result in an error unless you explicitly tell the program to perform the conversion because these risk losing information and causing other errors that are much more difficult to fix.

Implicit Conversions

Java can implicitly convert from any type earlier in this list to a type later in the list:

• byte
• short
• int
• long
• float
• double

Java will also implicitly convert from a char to int or any type below int on the list, but it will not implicitly convert any type to a char.

The following code demonstrates a few implicit conversions.

byte a = 10;
short b = a;
int c = b;
long d = c;
float e = d;
double f = e;

println(a + ", " + b + ", " + c + ", " + d + ", " + e + ", " + f);

Explicit Conversions

Any other conversions between numeric types must be explicitly stated in your code. If we don't do this, then we'll get a syntax error. If we reverse the order of the types in the last example, we'll see this happen with every line but the first. Try running the following script in jshell:

double a = 10;
float b = a;
long c = a;
int d = a;
short e = a;
byte f = a;

To make this work, we need to state the type we want to convert to in parenthesis before the expression we wish to convert:

double a = 10;
float b = (float)a;
long c = (long)a;
int d = (int)a;
short e = (short)a;
byte f = (byte)a;

println(a + ", " + b + ", " + c + ", " + d + ", " + e + ", " + f);