fastacf: Fast Autocorrelation
In phonTools: Tools for Phonetic and Acoustic Analyses

Description Usage Arguments Details Value Author(s) References Examples

Compute the ACF of a signal.

1 2	fastacf (signal, lag.max = length(signal), window = 'hann', show = TRUE, correct = FALSE)

`signal`	The signal, a numeric vector.
`lag.max`	The maximum lag value to be returned.
`window`	The type of window to be applied to the signal. Uses the windowfunc() function in this package. For no window select 'rectangular'.
`show`	If TRUE, the results are plotted.
`correct`	If TRUE, the output is corrected based on the window length and the window function that is applied.

The autocorrelation function is calculated using the inverse Fourier transform applied to the power spectrum. This leads to much faster calculation times than the acf() function included in the typical R installation. Corrections for window type and length are carried out as described in Boersma (1993).

A dataframe with the following columns:

`lag`	Indicates the lag value.
`acf`	Indicates the ACF value at the lag.

Santiago Barreda <[email protected]>

Boersma, P., (1993). Accurate short-term analysis of the fundamental frequency and the harmonics-to-noise ratio of a sampled sound. Proc. Instit. Phon. Sci. 17: 97-110.

## Uncomment and run the code below to see the speed advantage.
## Raising the n makes the difference even more pronounced.
#n = 25000
#system.time ({
#acf (rnorm (n), plot = F, lag.max = n)
#})

#system.time ({
#fastacf (rnorm (n), plot = F, lag.max = n)
#})