TimeSeriesDecompositionBase<T>

Base Classes · AiDotNet.DecompositionMethods.TimeSeriesDecomposition

Base class for time series decomposition algorithms that break down time series data into component parts.

Implements: ITimeSeriesDecomposition<T>

Known derived types: AdditiveDecomposition<T>, BeveridgeNelsonDecomposition<T>, EMDDecomposition<T>, HodrickPrescottDecomposition<T>, MultiplicativeDecomposition<T>, SEATSDecomposition<T>, SSADecomposition<T>, STLTimeSeriesDecomposition<T>, WaveletDecomposition<T>, X11Decomposition<T>

For Beginners

Time series decomposition is like taking apart a complex toy to see its individual pieces. It breaks down a sequence of data points (like daily sales, monthly temperatures, etc.) into simpler components that are easier to understand. Common components include:

  • Trend: The long-term direction (going up, down, or staying flat)
  • Seasonal: Regular patterns that repeat (like higher sales during holidays)
  • Residual: The leftover “noise” after removing trend and seasonal patterns

Properties

Property Summary
Vector<T> TimeSeries { get; } The original time series data to be decomposed.

Methods

Method Summary
Dictionary<DecompositionComponentType, object> GetComponents() Returns all components extracted from the time series.
Matrix<T> GetComponentAsMatrix(DecompositionComponentType componentType) Retrieves a specific component as a matrix.
Vector<T> GetComponentAsVector(DecompositionComponentType componentType) Retrieves a specific component as a vector.
bool HasComponent(DecompositionComponentType componentType) Checks if a specific component exists in the decomposition results.
object GetComponent(DecompositionComponentType componentType) Retrieves a specific component from the decomposition results.