### Autocorrelation function

`autocor [-v -p -P -o `*outfile*` -l# -x# -c# -V# -h] `* file(s)*

` -v `give unnormalised autocovariance

` -p `assume periodic continuation

` -P `assume periodic continuation exactly

` -l `number of values to be read (all)

` -x `number of values to be skipped (0)

` -c `column to be read (1 or *file*,#)

` -o `output file name, just ` -o `means *file*`_co`

` -V `verbosity level (0 = only fatal errors)

` -h `show this message

Computes the autocorrelation (or with `-v` the
autocovariance) function of a time series. This routine uses the FFT for
convolution and should thus asymptotically be faster than corr.
By default, zero padding is used, that is, no periodic continuation is assumed.
With the option `-P`, it is assumed that the
data is exactly one period of a periodic function. If the number of data points
is not factorisable with factors of 2,3, or 5, a slow Fourier transform would
have to be used. Therefore, with the option `-p`, approximate periodicity can be imposed by
finding the largest part of the time series that *is* thus factorisable.
Output file *file*`_co`.

See also corr

Table of Contents * TISEAN home