randEEG: EEG simulation

Description

Creates an object simulating ARIMA random variables. The created object contains data in the format required to use other methods of this package. Makes a simulation similar to EEG data to test the capabilities those methods. NOTE: The only purpose of the simulated data is to test the package features, it is not to be used to study properties of real EEG data!

Usage

randEEG(n.classes = 2, n.rec = 10, n.channels = 20, n.signals = 250, 
ar="default", ma="default", order="default",vars = c(1, 2))

Arguments

n.classes

number of different classes.

n.rec

number of recordings of each class.

n.channels

number of channels in the simulation.

n.signals

number of signals, observations or samples for each recording of each channel.

is a matrix with AR coefficients of the ARIMA model. Each row contains the AR parameters for each class and each channel. One can make the number of rows in ar equal to the number of classes and, in this case, the function uses the same coefficients for all channels of each class. Another option is to choose the number of rows equal to n.class*n.channels and, in this case, the simulation uses different ARIMA models for each class and each channel. For example if we have 2 classes and ar=matrix(c(0.1,0.5,-0.2,0.3,0,0),2,3,byrow=T) then the first class will be simulated from an AR(3) and the second class will be simulated from an AR(1) with parameter 0.3.

is a matrix with MA coefficients, defined the same way as ar. See the ar parameter for more information.

order

is a matrix with the same number of rows of ar and ma but with 3 columns. Thus, each line forms a vector which is used in the function arima.sim. For more information see the parameter order in the function arima.sim.

vars

a vector with length equal to the number of rows in ar, ma and order. Denote the variance for each ARIMA model.

Value

data: the simulated data frame. The data frame is organized as follows: each column represents a different channel. Thus, each signal collected by each channel is represented in each row. To identify the class and recording represented by each line, the vectors classes.Id and rec.Id are given and have length equal to the number of rows in the data frame.
classes.Id: the vector indicating the class id of each row of data.
rec.Id: the vector indicating the recording id of each row of data.
n.classes: number of different classes.
n.rec: number of recordings of each class.
n.channels: number of channels in the simulation.
n.signals: number of signals, observations or samples for each recording of each channel.
vars: a vector with length equal to the number of rows in ar, ma and order. Denote the variance for each ARIMA model.

References

Brockwell, P.J., Davis, R.A. (2002) Introduction to Time Series and Forecasting. 2nd ed. Colorado: Springer. Cap. 4.

Examples

Run this code

library(eegAnalysis)

#### Simulating EEG data
X<-randEEG(n.classes = 2, n.rec = 10, n.channels = 20, n.signals = 250, 
        vars = c(1, 2))


#### What do we have here?
names(X)


#### length of classes.Id and rec.Id are equal to the number of rows of data:
length(X$classes.Id)==nrow(X$data)
length(X$rec.Id)==nrow(X$data)


#### We have two classes:
unique(X$classes.Id)


#### We have 10 recordings for each class:
unique(X$rec.Id[which(X$classes.Id==1)])
unique(X$rec.Id[which(X$classes.Id==2)])


#### We have 20 channels:
ncol(X$data)


#### We have 250 signals for each class, recording and channel:
length(X$data[which(X$classes.Id==1 & X$rec.Id==1),1])