soundgen (version 2.9.0)

getSurprisal_vector: Get surprisal per vector

Description

Internal soundgen function called by getSurprisal. Estimates the unexpectedness or "surprisal" of the last element of input vector.

Usage

getSurprisal_vector(
  x,
  method = c("acf", "np", "none")[1],
  bestLag = NULL,
  weightByPrecision = TRUE,
  onlyPeakAutocor = FALSE
)

Arguments

x

numeric vector representing the time sequence of interest, eg amplitudes in a frequency bin over multiple STFT frames

method

(for $surprisal only, has no effect on $info and $kl) acf = change in maximum autocorrelation after adding the final point; np = nonlinear prediction (see nonlinPred - works but is VERY slow); none = do not calculate $surprisal to save time and only return $info and $kl

bestLag

(only for method = 'acf') if specified, we don't calculate the ACF but simply compare autocorrelation at bestLag with vs without the final point

weightByPrecision

if TRUE, surprisal is weighted by the current autocorrelation, so deviations from a previous pattern are more surprising if this pattern is strong

onlyPeakAutocor

if TRUE, only peaks of ACFs are considered (so bestLag can never be 1, and the first change after a string of static values results in surprisal = NA)

Examples

x = c(rep(1, 3), rep(0, 4), rep(1, 3), rep(0, 4), rep(1, 3), 0, 0)
soundgen:::getSurprisal_vector(x)
soundgen:::getSurprisal_vector(c(x, 1))
soundgen:::getSurprisal_vector(c(x, 13))

soundgen:::getSurprisal_vector(x, method = 'np')
soundgen:::getSurprisal_vector(c(x, 1), method = 'np')
soundgen:::getSurprisal_vector(c(x, 13), method = 'np')