Numeric types and operators reference¶

A compact lookup for Python's numeric types, the operators that work on them, and the numeric built-ins.

The types¶

Type	Literal	Exact?	Use for
`int`	`42`, `0xff`, `1_000`	yes, unbounded	counts, indices, ids, integer maths
`float`	`3.14`, `1e6`	no — IEEE 754 double	measurements, science, anything continuous
`complex`	`2 + 3j`	no (two floats)	signal processing, maths
`Decimal`	`Decimal('1.50')`	yes, base-10	money, decimal amounts with set rounding
`Fraction`	`Fraction(1, 3)`	yes, rational	exact ratios
`bool`	`True`, `False`	— (subclass of `int`)	truth values; `True == 1`

Decimal is in the decimal module, Fraction in fractions. Both should be built from strings, not floats.

Operator	Operation	Note
`+` `-` `*`	add, subtract, multiply
`/`	true division	always returns `float`
`//`	floor division	rounds toward −∞; `int` if both operands are `int`
`%`	remainder (modulo)	sign matches the divisor
`**`	exponentiation	`2 ** 0.5`; exact and unbounded for `int` exponents
`-x` `+x`	negation, unary plus

divmod(a, b) returns (a // b, a % b) together. Mixed-type expressions widen to the more general type (int + float → float).

Function	Returns
`abs(x)`	absolute value (or magnitude of a `complex`)
`round(x[, n])`	round half to even; `n` decimal places, same type back
`pow(x, y[, z])`	`x y`, or `(x y) % z` efficiently
`divmod(a, b)`	`(quotient, remainder)`
`min(...)` / `max(...)`	smallest / largest
`sum(iterable[, start])`	total (pass `Decimal('0')` to keep that type)
`int(x[, base])`	convert/parse to `int` (truncates floats toward zero)
`float(x)`	convert to `float`
`complex(re, im)`	build a complex number
`bin(n)` / `oct(n)` / `hex(n)`	base-prefixed string of an `int`

From → to	Behaviour	Example
`float` → `int`	truncates toward zero (not rounding)	`int(3.9)` → `3`, `int(-3.9)` → `-3`
`str` → `int`	integer strings only; `base` optional	`int('ff', 16)` → `255`
`str` → `float`	accepts decimals, `inf`, `nan`	`float('3.5')` → `3.5`
`str` → `int` (decimal text)	raises `ValueError`	use `int(float('3.5'))`
`int`/`str` → `Decimal`	exact	`Decimal('1.50')`
`float` → `Decimal`	exact but imports the float's error	avoid; prefer the string

Expression	Returns
`type(x)`	the exact type
`isinstance(x, int)`	type check (note `bool` is an `int`)
`n.bit_length()`	bits needed to represent `int` `n`
`x.is_integer()`	`True` if a `float` has no fractional part
`x.as_integer_ratio()`	`(numerator, denominator)` of a `float`/`Decimal`