filter.intensity: Filter a time-series vector/matrix based on its spectral...

In ben-oneill/ts.extend: Stationary Gaussian ARMA Processes and Other Time-Series Utilities

Filter a time-series vector/matrix based on its spectral intensity

Description

filter.intensity returns a vector/matrix of filtered time-series

Usage

filter.intensity(x, centered = TRUE, centred = centered, scaled = TRUE,
         highpass = NULL, highpass.remove = TRUE,
         lowpass  = NULL, lowpass.remove  = TRUE)

Arguments

x	A vector of time-series values
centered	Logical; if TRUE the time-series vector is centred (mean removed) before computing its intensity
centred	An alternative name for the centered input
scaled	Logical; if TRUE the intensity measure is scaled so that its norm is equal to the number of values in the time-series
highpass	The minimum intensity value for the highpass filter (or NULL if there is no highpass filter)
highpass.remove	Logical; if TRUE then the highpass filter removes (instead of attenuating) the low-frequency signal
lowpass	The maximum intensity value for the lowpass filter (or NULL if there is no lowpass filter)
lowpass.remove	Logical; if TRUE then the lowpass filter removes (instead of attenuating) the high-frequency signal

Details

This function filters a time-series based on its spectral intensity. The user inputs the time-series vector or matrix and specifies whether it is to be centered and/or scaled. Centering subtracts the sample mean of the vector prior to conversion into frequency space; this sets the intensity of the signal to zero at the zero frequency. Scaling scales the intensity so the that norm of the intensity vector is equal to the number of values in the time-series. The function then applies the specified highpass/lowpass filter and returns the filtered signals. By default, both the highpass and lowpass filters remove parts of the signal that do not meet the filter requirements; if preferred the filters can attenuate (rather than removing) these signals.

ben-oneill/ts.extend documentation built on May 4, 2023, 1:50 a.m.