unxt-hypothesis

Hypothesis strategies for unxt - unitful quantities in JAX.

This package provides Hypothesis strategies for generating unxt objects for for property-based testing.

Installation

uv add unxt-hypothesis

Usage

from hypothesis import given

import unxt_hypothesis as ust


@given(q=ust.quantities(unit="km/s"))
def test_quantity_property(q):
    # Test some property of quantities
    assert q.value is not None
    assert q.unit is not None


@given(u=ust.units("length"))
def test_unit_property(u):
    # Test some property of units
    assert u is not None


@given(sys=ust.unitsystems("m", "s", "kg", "rad"))
def test_unitsystem_property(sys):
    # Test some property of unit systems
    assert len(sys) == 4

Custom strategies

You can customize the strategies:

from hypothesis import strategies as st

import unxt as u
import unxt_hypothesis as ust


# Generate quantities with specific shapes
@given(q=ust.quantities(unit="m", shape=st.just((3, 3))))
def test_matrix_quantity(q):
    assert q.shape == (3, 3)


# Generate quantities with specific dimensions
@given(q=ust.quantities(unit=ust.units("length")))
def test_length_quantity(q):
    assert u.dimension_of(q) == u.dimension("length")

API

quantities(draw, *, shape=None, dtype=None, unit=None)

Generate random Quantity objects.

Parameters:

• draw: Hypothesis draw function
• shape: Strategy for array shapes (optional, defaults to small arrays)
• dtype: Strategy for array dtypes (optional, defaults to float32)
• unit: Strategy for unit strings (optional, defaults to common units)

Returns: A unxt.Quantity instance

units(draw, dimension=None, *, max_complexity=2, allow_non_integer_powers=False)

Generate random Unit objects.

Parameters:

• draw: Hypothesis draw function
• dimension: Physical dimension (optional, e.g., "length", "velocity")
• max_complexity: Maximum complexity of compound units (default: 2)
• allow_non_integer_powers: Allow non-integer powers (default: False)

Returns: A unxt.AbstractUnit instance

unitsystems(*units)

Generate random UnitSystem objects.

Parameters:

• *units: Variable number of unit specifications. Each can be:
  • str: Fixed unit string (e.g., "m", "kg")
  • unxt.AbstractUnit: Fixed unit object
  • Strategy: A Hypothesis strategy that generates units

Returns: A unxt.AbstractUnitSystem instance

Example:

from hypothesis import given

import unxt_hypothesis as ust


# Fixed unit system
@given(sys=ust.unitsystems("m", "s", "kg", "rad"))
def test_mks_system(sys):
    assert len(sys) == 4


# Varying length unit, other units fixed
@given(sys=ust.unitsystems(ust.units("length"), "s", "kg", "rad"))
def test_varying_length(sys):
    assert len(sys) == 4

Documentation

For comprehensive documentation, examples, and guides, see the unxt documentation.

License

BSD 3-Clause License. See LICENSE for details.

Contributing

Contributions are welcome! Please see the main unxt repository for contributing guidelines.