Skip to content

numpy-pydantic-types is a Python library providing numpy scalar data types compatible with pydantic, allowing for easy validation and enforcement of numerical precision in pydantic models.

License

Notifications You must be signed in to change notification settings

eksavazquez/numpy-pydantic-types

Repository files navigation

numpy-pydantic-types

numpy-pydantic-types is a Python library providing numpy scalar data types compatible with pydantic, allowing for easy validation and enforcement of numerical precision in pydantic models.

Overview

pydantic models rely on Python's native types (e.g., int, float, str) for data validation, which can be limiting when working with scientific or numerical applications that require specific numpy types, like float32, uint16, or int8. This library bridges that gap by introducing numpy scalar types as valid pydantic field types, allowing you to define precise, controlled models that integrate with scientific workflows.

Features

  • Numpy Scalar Compatibility: Define pydantic models using numpy scalar types like float32, uint32, int8, etc.
  • Strict Type Enforcement: Ensures model fields adhere to specific precision and range constraints.
  • Simple Integration: Easily import and use numpy types in your pydantic models.

Installation

pip install numpy_pydantic_types

Usage

This library enables the use of numpy scalar types directly within pydantic models. Here’s how to set up a model using numpy types:

from pydantic import BaseModel, ValidationError
from numpy_pydantic_types import Float32, UInt32, UInt16

class ScientificModel(BaseModel):
    precision_value: Float32  # Enforces a 32-bit floating point
    sample_count: UInt32  # Enforces a 32-bit unsigned integer
    sensor_id: UInt16  # Enforces a 16-bit unsigned integer

# Example data
data = {
    "precision_value": 1.234567,
    "sample_count": 4294967295,  # Max value for UInt32
    "sensor_id": 65535  # Max value for UInt16
}

# Instantiate the model with the specific numpy types
try:
    model = ScientificModel(**data)
    print("Model validated successfully:", model)
except ValidationError as e:
    print("Validation error:", e)

Supported Numpy Types

The library provides a range of numpy scalar types that can be used in pydantic models, including:

  • Float Types: Float32, Float64
  • Signed Integer Types: Int8, Int16, Int32, Int64
  • Unsigned Integer Types: UInt8, UInt16, UInt32, UInt64

Example Model

Using specific numpy scalar types can help enforce type constraints in applications where precision or memory footprint is crucial, such as scientific computations, data analysis, or embedded systems.

from numpy_pydantic_types import Int8, Float64

class DataProcessingModel(BaseModel):
    temperature: Float64
    adjustment_factor: Int8

Why Use Numpy Scalar Types?

  • Precision: Control over numerical precision, essential in scientific or numerical applications.
  • Range Enforcement: Ensures values conform to specific data type ranges (e.g., UInt8 ranges from 0 to 255).
  • Memory Efficiency: Helps reduce memory consumption by enforcing smaller data types.

License

MIT License

About

numpy-pydantic-types is a Python library providing numpy scalar data types compatible with pydantic, allowing for easy validation and enforcement of numerical precision in pydantic models.

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages