- Rust 97.9%
- C++ 2.1%
|
|
||
|---|---|---|
| examples | Update library name | |
| include | First successful C++ bindings! | |
| src | Update library name | |
| .gitignore | Update gitignore | |
| build.rs | First successful C++ bindings! | |
| Cargo.toml | New minor version due to switch to Rust 2024 | |
| LICENSE | Initialize community edition w/ MIT license | |
| README.md | Minor tweak to README | |
Elara Array (Community Edition)
This crate provides a minimalist, fast, n-dimensional array for Project Elara with minimal dependencies. Note: this crate was once part of elara-math, but has been extracted into a standalone crate for separate development.
Important information
There are two versions of Elara Array. Elara Array's original edition is public domain, but developed slowly. This repository houses Elara Array's community edition, which is licensed under a highly-permissive MIT license and is developed at a much more rapid pace. As such, the community edition is the recommended repository for new features requests and PRs.
Basic usage
useelara_array_community::prelude::*;fn main(){leta=arr![[1.0,2.0],[3.0,4.0]];println!("{:?}",a);}Design goals
elara-array is designed to be a minimalist library focused on a high-performance implementation of N-dimensional arrays on the CPU. GPU-accelerated arrays (tensors) are not within the scope of elara-array, but since the array implementation is very transparent, you are welcome to write your own code to transfer/copy the data from elara-array's arrays to the GPU.