#time-series #course #module #machine-learning #timer #model #built

timers-rs

Time series library for analysts, ML engineers, and data scientists. This package was actually built as a course module.

4 releases

0.1.3 Feb 6, 2025
0.1.2 Feb 6, 2025
0.1.1 Feb 4, 2025
0.1.0 Feb 4, 2025

#300 in Machine learning

Download history 116/week @ 2025-01-30 206/week @ 2025-02-06 7/week @ 2025-02-13 3/week @ 2025-02-27

235 downloads per month

MIT/Apache

24KB
434 lines

This package was actually built as a course module. In the course, I cover the math behind each time series model and build everything from scratch without using any external crates. Check out this course here: https://www.udemy.com/course/build-an-open-source-time-series-lib-from-scratch-in-rust/?referralCode=802A036F5A6D7453BCBD

Library Modules:

arima arma autocorrelation autoregressive sarima seasonality ses sma stationarity trend whitenoise wma

Sample Usage:

use timers_rs::autocorrelation::autocorrelation;
use timers_rs::ses;

fn main() {

    let series = vec![10.0, 12.0, 14.0, 16.0, 18.0, 20.0];

    let model = ses::ExponentialSmoothing{
        alpha: 0.5,
        beta: Some(0.0),
    };

    let single_exponential_smoothing = model.single_exponential_smoothing(&series);
    println!("Single Exponential Smoothing: {:?}", single_exponential_smoothing);

    let autocorrelation_result = autocorrelation(&series, 2);
    println!("Autocorrealtion Result {}", autocorrelation_result);

}

No runtime deps