boundednumbers

Boundednumbers provides lightweight numeric types that automatically respect the limits you set. Use them to keep values in range across math operations without littering your code with manual clamping or wraparound logic.

Installation

pip install boundednumbers

Quick start

from numbers import ClampedInt, CyclicInt, BouncedInt, ModuloInt, UnitFloat, modulo_range

health = ClampedInt(120, 0, 100)   # -> 100, and stays within [0, 100]
angle = CyclicInt(-15, 0, 360)     # -> 345, wraps on every operation
paddle = BouncedInt(130, 0, 100)   # -> 70, reflects back into range
counter = ModuloInt(7, 5)          # -> 2, supports modular arithmetic
opacity = UnitFloat(1.25)          # -> 1.0, always between 0 and 1

# Iterate around a circle without repeating the stop value
points = list(modulo_range(start=350, stop=170, step=40, modulus=360))

Features

• Bounded integers: ClampedInt, CyclicInt, BouncedInt, and ModuloBoundedInt automatically enforce their strategy after arithmetic (addition, subtraction, multiplication, division, and floor division).
• Bounded floats: ClampedFloat, CyclicFloat, and BouncedFloat mirror the integer behavior for floating-point values.
• Unit interval helpers: UnitFloat corrects values into [0, 1] on construction; EnforcedUnitFloat reapplies the bounds after arithmetic.
• Modular arithmetic: ModuloInt exposes additive and multiplicative inverses along with safe modular iteration via modulo_range.
• Utility functions: Reusable helpers like clamp, clamp01, bounce, cyclic_wrap, and cyclic_wrap_float are available when you prefer to keep native numeric types.

API overview

Bounded factories

Use the factory helpers when you need custom class names or number types:

from numbers import BoundType, bounce, make_default_bounded_int, make_default_bounded_float, make_bounded_int

# Create a custom bounded integer that bounces inside [-5, 5]
BouncyFive = make_bounded_int(lambda v, mn, mx: bounce(v, mn, mx), class_name="BouncyFive")
value = BouncyFive(17, -5, 5)  # -> -3

# Or rely on the built-in enum to pick a strategy
WrappedFloat = make_default_bounded_float(BoundType.CYCLIC)
angle = WrappedFloat(725.0, 0.0, 360.0)  # -> 5.0

Modular utilities

ModuloInt and modulo_range simplify common modular workflows:

from numbers import Direction, ModuloRangeMode, modulo_range, ModuloInt

# Compute a multiplicative inverse
inverse = ModuloInt(7, 26).inverse()  # -> ModuloInt(15 mod 26)

# Visit points around a circle without repeating stop (one full modular cycle)
for value in modulo_range(start=0, stop=0, step=3, modulus=10, direction=Direction.DECREASING, max_range_amount=ModuloRangeMode.DETECT):
    ...

Unit floats

UnitFloat corrects values into the unit interval on construction, while EnforcedUnitFloat applies bounds after every operation:

from numbers import EnforcedUnitFloat

u = EnforcedUnitFloat(1.4)
print(u - 3.0)  # -> 0.0

Further reading

See numbers/README.md for a deeper, module-by-module breakdown and usage tips.