Numeric expressions in BASIC-256 come in two varieties; 1) integer, and 2) floating point numbers. They are created by using Arithmetic Operators and functions with Numeric Constants and Variables.
Integer Expressions are expressions that return a whole number.
Examples include:
Integer Expressions |
---|
10 |
ABS(-2345) |
int(9.987) |
5*8 |
Positive integer values may also be expressed in binary (base 2), octal (base 8), and hexadecimal (base 16). Precede binary values with 0b (0b1110 = 14), octal with 0o (0o177 = 127), and hexadecimal with 0x (0xff = 255).
Expressions that return a decimal number, even if it ends in .0, are known as floating point expressions.
Examples include:
Floating Point Expressions |
---|
3/7 |
float(56) |
sin(2*pi) |
Very large and small floating point numbers may also be represented in scientific E notation (added in version 0.9.9.46). http://en.wikipedia.org/wiki/Scientific_notation
Scientific E Notation | |
---|---|
E Notation | Decimal Number |
2e0 | 2 |
3e2 | 300 |
1.234e10 | 12340000000 |
-5.3e4 | -53000 |
2e-1 | 0.2 |
5.12e-9 | 0.00000000512 |