rustkernel-temporal
GPU-accelerated temporal analysis kernels for forecasting, decomposition, detection, and volatility modeling.
Kernels (7)
Forecasting (2 kernels)
- ARIMAForecast - ARIMA(p,d,q) model fitting and forecasting
- ProphetDecomposition - Prophet-style trend/seasonal/holiday decomposition
Detection (2 kernels)
- ChangePointDetection - PELT/Binary segmentation/CUSUM
- TimeSeriesAnomalyDetection - Statistical threshold detection
Volatility (2 kernels)
- GARCHVolatility - GARCH(1,1) volatility modeling
- RealizedVolatility - High-frequency realized volatility
Seasonality (1 kernel)
- SeasonalDecomposition - STL decomposition
Installation
Add to your Cargo.toml:
[]
= "0.1.0"
Usage
use *;
// Forecast with ARIMA
let arima = new;
let forecast = arima.fit_and_forecast;
License
Apache-2.0