noise: Noise

In connolr3/eegdatasim: Tool for EEG signals and ERPs

Noise

Description

Generates Classicaly simulated Noise from an EEG signal. Adapted from Matlab version. Functioanlity is the same, however it avoids O(N^2) time complexity of Matlab version, as R is not suited well to for loops. Uses apply and matrices instead. meanpower is read in at start in noise function, if not supplied by user.

Usage

noise(x)

Arguments

frames	number of signal frames per each trial
epochs	number of simulated trials
srate	sampling rate of simulated signal in Hz

Details

The value of the first parameters describing the number of samples is computed by multiplying the duration of the noise by the sampling frequency, i.e. if we wanted 0.8 seconds of noise, with a srate of 250Hz.... choose frames = 200 0.8 * 250 = 200 The function generates a vector containing the samples.

Noise is generated such that its power spectrum matches the power spectrum of human EEG

Value

signal

A vector of length frames*Epochs, representing (epochs) number of trials, each of length (frames), concatenated into one vector.

Author(s)

Adapted from Yeung N,Bogacz R, Holroyd C, Nieuwenhuis S, Cohen J

References

https://data.mrc.ox.ac.uk/data-set/simulated-eeg-data-generator

Examples

fill_signals<-function(mysignal,frames, epochs, srate,meanpower){
  sumsig = 50#number of sinusoids from which each simulated signal is composed of
  signal <- matrix( rep(1:frames, sumsig), nrow=sumsig, byrow=TRUE )
  freq <- 4 * runif( sumsig, 0, 1)#generate random frequency for each sin
  freq <- cumsum(freq)#apply cumilitive function to make each sin have a higher frequency than the last
  freqamp <- meanpower[ pmin( ceiling(freq), rep(125,sumsig)) ] / meanpower[1]#generate ampltidue based on meanpower
  phase <- 2 * pi * runif(sumsig, 0 , 1)#generate random phase for each sin
  signal <- sin( signal * freq * ( 2 * pi / srate ) + phase ) * freqamp#create 50 sins, with consecutively higher freq for each sin 
  mysignal<-colSums( signal ) 
  
}

noise <- function(frames, epochs, srate, meanpower = NULL) {
  if (frames < 0) stop("frames cannot be less than 0")
  if (srate < 0) stop("srate cannot be less than 0")
  if (epochs < 0) stop("epochs cannot be less than 0")
  if( is.null(meanpower)) meanpower <- as.vector(R.matlab::readMat("meanpower.mat")$meanpower)
  signals<-matrix(0,frames,epochs)
  signals<-apply(signals,2,fill_signals,meanpower=meanpower,frames=frames,epochs=epochs,srate=srate)
  return(as.vector(signals))
}


#mynoise<-noise(200,10,250)
#plot(mynoise[1:200])

connolr3/eegdatasim documentation built on April 14, 2022, 10:39 a.m.