dot-fps_fpca: Internal FPCA helper
In FPScausal: Functional Propensity Score for Causal Inference

.fps_fpca

R Documentation

Internal FPCA helper

Description

Performs Functional Principal Component Analysis on a matrix of functional observations using a B-spline basis expansion, then retains enough components to explain at least 'pve' of the total variance. Accepts either an n x T numeric matrix or an fd object from the fda package.

Usage

.fps_fpca(X, pve = 0.95, t_grid = NULL, domain = NULL, nbasis = NULL)

Arguments

`X`	n x T numeric matrix of observed functional data (rows = subjects), or an `fd` object.
`pve`	Proportion of variance explained threshold (default 0.95).
`t_grid`	Numeric vector of observation time points (length T). Inferred from `X` when `X` is an `fd` object.
`domain`	Numeric vector c(a, b) giving the domain of the functions. Inferred from `X` when `X` is an `fd` object.
`nbasis`	Number of B-spline basis functions. If NULL, set automatically.

Value

A named list with components:

`scr`	n x L matrix of FPC scores.
`efn`	T x L matrix of eigenfunction values evaluated on t_grid.
`mean`	Numeric vector of length T: mean function evaluated on t_grid.
`eval`	Numeric vector of eigenvalues (all, not just L).
`varprop`	Numeric vector of variance proportions for retained components.
`perc`	Cumulative variance proportions (all components).
`pca_fd`	The raw `pca.fd` object from the fda package.
`L`	Number of retained components.
`t_grid`	The t_grid used.
`domain`	The domain used.