Skip to contents

Elastic (Karcher) mean of a set of curves

Source: R/align-elastic.R
srvfMean.Rd

Computes the elastic template mean of repeated curves by iteratively aligning every curve to the current mean and averaging the aligned curves. Because the phase variability is removed before averaging, peaks are preserved rather than blurred out (unlike a naive cross-curve mean).

Usage

srvfMean(curves, t = NULL, iterations = 5L, max_step = 6L, use_fdasrvf = NA)

Arguments

curves

A list of numeric curves of the same length, or a matrix whose columns are curves.

t

Optional grid (defaults to 0..1).

iterations

Number of refinement iterations (default: 5).

max_step

Maximum DP step for the pure-R solver (default: 6).

use_fdasrvf

As in elasticAlign().

Value

A list with the elastic mean curve, the warpings (a matrix, one column per curve), and the aligned curves.

See also

elasticAlign()

Examples

t <- seq(0, 1, length.out = 80)
curves <- lapply(c(-0.1, 0, 0.1), function(s)
  exp(-((t - (0.5 + s))^2) / (2 * 0.05^2)))
m <- srvfMean(curves)
max(m$mean)
#> [1] 0.991812