numpy-lite 1.26.4

Ultra-light minimal NumPy build (core functionality only) for AWS Lambda / serverless.

numpy-lite

Ultra-light minimal NumPy 1.26.4 build focused on core functionality only.  

This fork is designed for AWS Lambda / serverless environments with Python 3.12, removing heavy modules like BLAS, LAPACK, linalg, fft, polynomial, random, and f2py to reduce package size.

Features

• Core array operations (np.array, np.sum, np.mean, indexing, slicing)
• Minimal footprint (~8MB)
• AWS Lambda / serverless ready
• Stub modules provide informative errors if removed modules are imported

Installation

Install via PYPI (recommended)

The easiest way to install:

pip install numpy-lite

###Install from GitHub (for development or contributions)

Clone or download the repository:

git clone https://github.com/JacquieAM/numpy-lite.git
cd numpy-lite

Install via pip:

pip install .
rm -rf build

Usage

import numpy as np

arr = np.array([1, 2, 3])
print(arr)
print(np.sum(arr))

Modules like np.linalg, np.fft, and np.random will raise informative ImportError if used.