tsdownsample
Panther
Our great sponsors
tsdownsample | Panther | |
---|---|---|
7 | 2 | |
127 | 213 | |
6.3% | - | |
6.0 | 4.3 | |
16 days ago | over 1 year ago | |
Jupyter Notebook | Rust | |
MIT License | MIT License |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
tsdownsample
- downsampling 500M datapoints in < 0.05s
-
[P] tsdownsample: extremely fast time series downsampling for visualization
P.S.: I recently discovered a bug in the implementation (when there are large gaps in the time series ) -> PR https://github.com/predict-idlab/tsdownsample/pull/20 should fix this
- tsdownsample: Timeseries downsampling with 200-300x better performance vs. NumPy
- tsdownsample: time-series downsampling with 200-300x performance to equivalent numpy routines
-
tsdownsample: extremely fast time series downsampling written in Rust
I just created an Issue for this: https://github.com/predict-idlab/tsdownsample/issues/19
Panther
- Panther: A high performance Python technical analysis library written in Rust using PyO3 and rust-numpy. 9x faster than pandas alone!
-
Man, I love this language.
I recently started learning rust and decided to make a python library with PyO3 and NDArray as a first project. With the name Panther, the library was supposed to be an implementation of stock technical indicators (EMA, SMA, RSI, Ect). I added a few functions, and decided to do some speed tests with the pandas way of calculating these indicators. I was shocked to see that my code was about 9x faster on average than pandas calculations. I know this is expected when using a low level language like rust in python, but I'm amazed none the less. Especially as someone new to rust, the fact I could get these "advertised" results with rust in python without having to do crazy optimizations is crazy to me. Plus, something about writing low-level code and getting these results in python is very satisfying. The best part though? The process to get these results wasn't even hard! The cargo packages I used had great documentation and the compiler?! Actually helpful! With reference material on errors too! Officially done geeking out about Rust haha but love this language and love this community. Hoping to get more involved with OS stuff. What project is everyone working on? Anything cool?
What are some alternatives?
ZenithTA - A high performance python technical analysis library written in Rust and the Numpy C API.
boing - A safe wrapper over libui-ng-sys.
tsai - Time series Timeseries Deep Learning Machine Learning Pytorch fastai | State-of-the-art Deep Learning library for Time Series and Sequences in Pytorch / fastai
polars - Dataframes powered by a multithreaded, vectorized query engine, written in Rust
downsample - Collection of several downsampling methods for time series visualisation purposes.
lazerpay-rust-sdk - Lazerpay SDK for Rust 🦀
argminmax - Efficient argmin & argmax
tendie-factory - Tendie-Factory is a work in progress application that seeks to track the stocks mentioned in the wallstreetbets subreddit.
docs.rs - crates.io documentation generator
rust-numpy - PyO3-based Rust bindings of the NumPy C-API
tealr_doc_gen - an online documentation generator for apis written with tealr