Pint: Physical Quantities

0.25.3 · active · verified Thu Apr 09

Pint is a Python package designed for defining, operating, and manipulating physical quantities. It allows for arithmetic operations between numerical values and units of measurement, as well as conversions between different units. The library includes a comprehensive list of physical units, prefixes, and constants, and its modular design enables users to extend or rewrite unit definitions. It integrates well with numerical libraries like NumPy and Pandas (via `pint-pandas`) and is actively maintained, with the latest stable version 0.25.3 released on March 19, 2026.

Warnings

breaking Minimum Python version support has increased over time. For Pint v0.25.3, Python 3.11+ is required. Older versions like 0.18 required Python 3.8+ and Numpy 1.19+; version 0.10 required Python 3.6+. Check the changelog for specific version compatibility.
Fix: Ensure your Python environment meets the minimum version requirement for your chosen Pint release (e.g., Python >=3.11 for 0.25.3) and update NumPy if necessary.
gotcha Creating multiple instances of `UnitRegistry` across different modules in your project can lead to `ValueError: Cannot operate with Quantity and Quantity of different registries.` when performing operations between them.
Fix: Instantiate the `UnitRegistry` in a single, central location (e.g., your package's `__init__.py` file) and import that shared instance across your project. Optionally, use `set_application_registry(ureg)` for advanced scenarios.
breaking Attempting to cast a `Quantity` with offset units (e.g., temperature in Celsius) to a boolean value will raise a `ValueError` since version 0.10. This prevents ambiguous boolean evaluations.
Fix: Explicitly compare the quantity's magnitude to a reference value (e.g., `if temp.magnitude > 0:`) rather than relying on implicit boolean conversion.
gotcha The `auto_reduce_dimensions` and `autoconvert_to_preferred` parameters of `UnitRegistry` are `False` by default for performance reasons. This means units might not automatically simplify or convert to preferred forms unless explicitly enabled.
Fix: If automatic simplification or conversion is desired, initialize your `UnitRegistry` with `ureg = UnitRegistry(auto_reduce_dimensions=True, autoconvert_to_preferred=True)` or call `.to_reduced_units()` / `.to_preferred()` explicitly.
gotcha There is a separate Python library for pulsar timing called `PINT` (all uppercase). Be careful not to confuse `pint` (lowercase, for physical quantities) with `PINT` when installing or searching, as they are distinct projects.
Fix: Ensure you are installing `pint` (lowercase) for physical quantities and refer to its specific documentation and examples to avoid conflicts or incorrect usage.

Install

pip install pint PyPI
conda install -c conda-forge pint Conda

Imports

UnitRegistry
```
from pint import UnitRegistry
```

Quickstart

Initialize a `UnitRegistry` to define and manage units. Use the registry to create `Quantity` objects, perform arithmetic operations, and convert units. Pint automatically handles dimensional analysis during calculations.

from pint import UnitRegistry

# Initialize the UnitRegistry, which stores unit definitions and handles conversions
ureg = UnitRegistry()

# Define quantities using numerical values and units from the registry
distance = 24.0 * ureg.meter
time = 8.0 * ureg.second

# Perform calculations; Pint automatically handles unit propagation
speed = distance / time
print(f"Calculated Speed: {speed}")

# Convert quantities to different compatible units
speed_kmh = speed.to(ureg.kilometer / ureg.hour)
print(f"Speed in kilometers per hour: {speed_kmh}")

# Quantities can also be created by parsing strings
force = ureg.Quantity("100 N")
print(f"Force from string: {force}")

view raw JSON →