8 releases
0.2.6 | Jul 6, 2024 |
---|---|
0.2.5 | Jul 3, 2024 |
0.2.4 | Jun 2, 2024 |
0.2.3 | May 30, 2024 |
0.1.0 | Sep 2, 2023 |
#210 in Math
254 downloads per month
200KB
5.5K
SLoC
twibint
Rust crate for arbitrarily big integers, signed or unsigned.
This crate does not need any dependency, and relies only on the standard library. Some dependencies are optional, depending on a specific feature (see list of features below).
The main API of this crate is to export 2 types: BigUint and BigInt,
meant to represent unsigned or signed integers of arbitrarily large
absolute value. They are meant to be used in almost any way a regular
integer can be used, though they don't implement the Copy
trait.
Build, documentation, benchmarks and tests are available the usual way calling the following:
cargo build
cargo docs
cargo bench --features=rand
cargo test
For benchmarks specifically, you might want to call only some of these:
cargo bench mul --features=rand
cargo bench add --features=rand
cargo bench sub --features=rand
Benchmarks won't compile/run without the rand
feature enabled.
Performance
The ambitious and naive goal of this project is to be as performant or better than any state of the art crate on these methods.
I choose to compare myself to num-bigint
first, as it's quite standard at this
point.
Today, on x86, twibint
is faster than num-bigint
v0.4 for addition, above
around 10000 bits. It is on par for multiplication, starting 1000 bits.
List of features
rand
: enables the possibility to generate a random integer with a specific number of bits. Usesrand
crate as a dependency.pyo3
: Only used to generate python bindings, it's only meant to be used indirectly via thepip install .
command. Usespyo3
crate as a dependency.unsafe
: Enables accelerations that use unsafe Rust. Enabled by default. Disabled via the flag--no-default-features
for pure safe Rust (compile time enforced).
Install as a Python package
Simply use from the base directory
python3 -m pip install .
This crate seems faster than the default Python integers for addition and multiplication above a certain numbers of bits (between 1000 and 10000 bits).
Python tests are available to be run in the pytest
framework. They are located
in the tests
folder and should provide ample example usage.
Changelog for version 0.2
This new version contains extensive accelerations for addition, subtraction, and multiplication on x86_64 machines. I used no modern extensions of x86, so these acceleration should be portable accross this family of mahcines. These will probably also have performance repercussions on many other features.
These acceleration are mostly due to dropping inline assembly for core loops, and are
based on unsafe
Rust. Other unsafe
features used include smartly swapping between
&[u32]
and &[u64]
slices via pointers (when alignment is lucky).
To disable any unsafe
code, use with the flag --no-default-features
for a slower
experience, but fully compiled in safe Rust.
Dependencies
~0–5.5MB
~13K SLoC