Skip to contents

Definition function: Lag two datasets, so as to maximise their cross-correlation

def.lag2.Rd

Function definition. Lag two datasets, so as to maximise their cross-correlation. Modified to expose freqThsh (cutoff frequency for high pass filter as an argument)

Usage

def.lag2(
  refe,
  meas,
  freq,
  freqThsh = (1/(2 * lagMax/freq)),
  dataRefe = refe,
  dataMeas = meas,
  measVar = NULL,
  lagMax = 2 * freq,
  lagCnst = TRUE,
  lagNgtvPstv = c("n", "p", "np")[3],
  lagAll = TRUE,
  hpf = TRUE,
  fracMin = 0.1
)

Arguments

refe

A vector with variable in reference time frame. Of class numeric. (-)

meas

A vector with variable in time frame to be adjusted. Of class numeric. (-)

freq

Acquisition frequency of refe and meas. Of class ingeter. (Hz)

freqThsh

cutoff frequency for the low end of the high pass filter. Defaults to 1 / (2 * lagMax / freq), the same as eddy4r.base::def.lag() (Hz)

dataRefe

A matrix or data.frame with all data that carries the time frame of refe. Defaults to refe. Of any class. (-)

dataMeas

A matrix or data.frame with all data that carries the time frame of meas. Defaults to meas. Of any class. (-)

measVar

A vector specifying if only several columns in dataMeas shall be lagged. Defaults to NULL. Of class integer or character. (-)

lagMax

Maximum lag, by default 2 x freq. Of class integer. (-)

lagCnst

TRUE - interpret lagMax as maximum permissible lag; FALSE - start with lagMax as first estimate and increase iteratively. Defaults to TRUE. Of class logical. (-)

lagNgtvPstv

"n" - consider negative lag times only, i.e. meas is expected to lag behind refe; "p" - consider positive lag times only, i.e. refe is expected to lag behind meas; "np" - consider negative and positive lag times. Defaults to "np". Of class character. (-)

lagAll

TRUE - consider positive and negative correlations when finding lag time; FALSE - consider positive correlations only when finding lag time. Defaults to TRUE. Of class logical. (-)

hpf

TRUE - apply Butterworth high-pass filter; FALSE - use raw data. Defaults to TRUE. Of class logical. (-)

fracMin

Minimum fraction of data to attempt lag determination. Defaults to 0.1. Of class numeric. (-)

Value

Lagged input data and calculation results in a list consisting of:

dataRefe The reference data. dataMeas The data that was lagged to coincide with the reference data. lag The number of data points by which the lag correction was performed. corrCros The cross-correction coefficient between refe and meas for the determined lag time.

References

License: GNU AFFERO GENERAL PUBLIC LICENSE Version 3, 19 November 2007.

Author

Stefan Metzger eddy4R.info@gmail.com