Crate so_tsa 

Time Series Analysis (TSA) module for StatOxide

This crate provides comprehensive time series analysis tools, including:

1. Core Data Structures: TimeSeries with datetime indexing
2. Stationarity Tests: ADF, KPSS, PP tests
3. ARIMA Models: AR, MA, ARMA, ARIMA, SARIMA
4. GARCH Models: ARCH, GARCH for volatility modeling
5. State Space Models: Kalman filter, structural time series
6. Decomposition: Seasonal-Trend decomposition (STL), moving averages
7. Forecasting: Point forecasts, prediction intervals
8. Diagnostics: Residual analysis, model selection criteria

Example Usage

use so_tsa::{TimeSeries, ARIMA, GARCH, GARCHDistribution};
use ndarray::Array1;

// Create a simple time series with enough data for ARIMA
let values = Array1::from_vec(vec![1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0, 12.0, 13.0, 14.0, 15.0]);
let timestamps: Vec<i64> = (0..values.len() as i64).collect();
let ts = TimeSeries::new("series", timestamps, values, None).unwrap();

// Fit ARIMA(0,0,0) model (white noise with constant)
let arima = ARIMA::builder(0, 0, 0)
    .with_constant(true)
    .max_iter(200)
    .tol(1e-4)
    .fit(&ts).unwrap();

// Fit GARCH(1,1) model  
let garch = GARCH::builder(1, 1)
    .distribution(GARCHDistribution::Normal)
    .max_iter(200)
    .tol(1e-4)
    .fit(&ts).unwrap();

References

• Box, G. E. P., Jenkins, G. M., Reinsel, G. C., & Ljung, G. M. (2015). Time Series Analysis.
• Hyndman, R. J., & Athanasopoulos, G. (2021). Forecasting: Principles and Practice.
• R’s forecast package and statsmodels’ tsa module.

Re-exports

pub use arima::ARIMA;

pub use arima::ARIMAResults;

pub use arima::SARIMABuilder;

pub use arima::SARIMAOrder;

pub use decomposition::DecompositionExt;

pub use decomposition::DecompositionMethod;

pub use decomposition::DecompositionResults;

pub use decomposition::HodrickPrescottFilter;

pub use decomposition::MovingAverageDecomposition;

pub use decomposition::STLDecomposition;

pub use decomposition::X12ARIMA;

pub use forecast::ForecastMetrics;

pub use forecast::IntervalMethod;

pub use forecast::PredictionInterval;

pub use forecast::PredictionIntervals;

pub use forecast::TimeSeriesCV;

pub use garch::ARCH;

pub use garch::GARCH;

pub use garch::GARCHDistribution;

pub use garch::GARCHOrder;

pub use garch::GARCHResults;

pub use results::ModelComparison;

pub use results::ResidualDiagnostics;

pub use results::TSAResults;

pub use statespace::KalmanFilter;

pub use statespace::StateSpaceModel;

pub use stationarity::ADFTest;

pub use stationarity::KPSSTest;

pub use stationarity::PPTest;

pub use stationarity::StationarityTest;

pub use timeseries::TimeSeries;

pub use utils::acf;

pub use utils::box_cox;

pub use utils::box_cox_lambda;

pub use utils::ccf;

pub use utils::detrend_poly;

pub use utils::diebold_mariano;

pub use utils::ewma;

pub use utils::forecast_errors;

pub use utils::information_criteria;

pub use utils::pacf;

pub use utils::periodogram;

pub use utils::rolling_statistic;

pub use utils::seasonal_dummies;

pub use utils::spectrum;

Modules

arima

ARIMA (AutoRegressive Integrated Moving Average) models

decomposition

Time series decomposition methods

forecast

Forecasting evaluation and prediction intervals

garch

GARCH (Generalized Autoregressive Conditional Heteroskedasticity) models

prelude

results

Unified result structures for time series analysis

statespace

State space models and Kalman filter

stationarity

Stationarity tests for time series

timeseries

Time series data structure with datetime indexing

utils

Utility functions for time series analysis