Reconstruct waveforms from fPCA

Reconstructs individual waveforms using a subset of principal components.

Usage

reconstructFPCA(fpca_result, n_components = NULL, observation = NULL)

Arguments

fpca_result

An fpca_result object from fPCA().

n_components

Number of components to use for reconstruction.

observation

Indices of observations to reconstruct. If NULL, all.

Value

Matrix of reconstructed waveforms (time x observations).

References

Ramsay JO, Silverman BW (2005). "Functional Data Analysis." 2nd ed. Springer.

See also

fPCA() for performing the decomposition, plotFPCA() (in PhysioMoCap) for visualizing fPCA results.

Examples

# Create sample data and run fPCA first
set.seed(123)
t <- seq(0, 100, length.out = 100)
base_curve <- sin(2 * pi * t / 100) * 30
data <- sapply(1:20, function(i) base_curve * rnorm(1, 1, 0.2) + rnorm(100, 0, 2))
fpca_result <- fPCA(data, n_components = 4)

# Reconstruct using only first 2 PCs
reconstructed <- reconstructFPCA(fpca_result, n_components = 2)