Module detrend 

Detrending and decomposition functions for non-stationary functional data.

This module provides methods for removing trends from functional data to enable more accurate seasonal analysis. It includes:

• Linear detrending (least squares)
• Polynomial detrending (QR decomposition)
• Differencing (first and second order)
• LOESS detrending (local polynomial regression)
• Spline detrending (P-splines)
• Automatic method selection via AIC

Structs

DecomposeResult

Result of seasonal decomposition.

StlResult

Result of STL decomposition including robustness weights.

TrendResult

Result of detrending operation.

Functions

auto_detrend

Automatically select the best detrending method using AIC.

decompose_additive

Additive seasonal decomposition: data = trend + seasonal + remainder

decompose_multiplicative

Multiplicative seasonal decomposition: data = trend * seasonal * remainder

detrend_diff

Remove trend by differencing.

detrend_linear

Remove linear trend from functional data using least squares.

detrend_loess

Remove trend using LOESS (local polynomial regression).

detrend_polynomial

Remove polynomial trend from functional data using QR decomposition.

stl_decompose

STL Decomposition: Seasonal and Trend decomposition using LOESS

stl_fdata

Wrapper function for functional data STL decomposition.