Skip to contents

Pointwise intraclass correlation across the movement cycle

Source: R/stats-reliability.R
waveformICC.Rd

The reliability curve: the between-subject ICC computed frame-by-frame from repeated trials, using the ecosystem's icc() at each cycle point. It shows where along the movement a measurement is reliable.

Usage

waveformICC(x, model = "twoway", type = "agreement", unit = "single")

Arguments

x

A [subject, trial, frame] array, or a list of trials x frames matrices (one per subject). Every subject must have the same trials and frames.

model, type, unit

Passed to icc() (default two-way, absolute agreement, single measure – ICC(2,1)).

Value

a waveform_icc list: icc, ci_lower, ci_upper (per frame), mean_icc, min_icc.

See also

waveformReliability(), waveformCMC()

Examples

set.seed(1)
subs <- lapply(1:8, function(s) {
  base <- s + sin(seq(0, 2 * pi, length.out = 101))   # subject offset = signal
  t(sapply(1:3, function(k) base + rnorm(101, 0, 0.1)))
})
waveformICC(subs)$mean_icc                            # high (subjects separable)
#> [1] 0.9982253