baselineCorr: Baseline correction
In tdeenes/eegR: Analyze EEG Signals

Description Usage Arguments Details Value See Also Examples

baselineCorr performs baseline correction, that is, it removes the average of the baseline amplitudes from the data.

baselineCorr(
  dat,
  pool_dim = "time",
  sep_dim = NULL,
  base_timerange = isPositive(negate. = TRUE),
  return_call = TRUE
)

`dat`	numeric vector, matrix, or array containing the ERPs
`pool_dim, sep_dim`	numeric or character vector identifying the dimensions of `dat` whose dimension levels should be pooled or treated separately, respectively. Provide only `pool_dim` or `sep_dim`, but not both (the default is: `pool_dim = "time"`). See Details and Examples.
`base_timerange`	numeric or character vector, or a subsetting function from the `is` family of functions identifying time points which form the baseline. The default is `isPositive(negate. = TRUE)`, which uses all time points up to zero. If NULL, all time points are included.
`return_call`	logical value whether the function call should be returned with additional informations (default: TRUE)

Baseline correction is a standard step in ERP preprocessing. It means the subtraction of the average prestimulus voltage from the whole waveform. Usually baseline correction is done separately for each channel in each epoch, that is, each single ERP curve is corrected independently. baselineCorr allows to provide non-standard time windows and the pooling of dimensions (e.g., a common baseline is computed for all conditions).

The function returns a numeric array with the same attributes as dat.

scaleArray for the function behind the scenes

# example dataset
data(erps)

# remove baseline activity - separately for each single time curve 
# (this is the default, but we explicitly include it for now) 
bc0 <- baselineCorr(erps, pool_dim = "time")

# remove baseline activity separately for each stimulus class, pairtype and 
# channel in each subject
bc1 <- baselineCorr(erps, sep_dim = c("stimclass", "pairtype", "chan", "id"))

# in the present case, the two calls give the same results, because
# the ERP dataset has these five dimensions
stopifnot(all.equal(bc0, bc1, check.attributes = FALSE)) 

# assume that the baseline is from -50 ms to 50 (silly example)
bc3 <- baselineCorr(erps, base_timerange = isBetween(-50, 50))