CreateAllSCM: Creation of sample covariance matrices

In LauraMasiero/RiemannR: RiemannR - EEG-based BCI analysis with Riemannian geometry

Description

This function creates the sample covariance matrices of multiple matrices created with or without the function SuperTrial.

Usage

createAllSCM(list, n=8, st=T, tidy=FALSE, col=17)

Arguments

list	a list of matrices with a specific construction (if st=F) or a list of matrices constructed by the function SuperTrial (if st=T). See Details below.
n	a numeric value indicating how many SCM have to be computed.
st	boolean. If st=T the elements of the list in input are matrices constructed with the function SuperTrial, if st=F the matrices have a particular construction, described in Details below.
tidy	option to remove specific channels. If tidy=TRUE no channels will be removed to construct the Super trial, if tidy=FALSE then the columns of the channels to be removed have to be specified with the option col.
col	columns to be removed (when tidy=FALSE).

Details

The initial list must have a specific construction: every list element has to be a list itself with two main elements, $epochs and $labels. $epochs contains tensors with dimensions [n,c,t], where n are the number of epochs, c are the number of channels and t is the recording (epoch) length. $labels contains the class of the epochs considered (i.e. target or non target).

Value

It returns a list containing the sample covariance matrices computed.

Author(s)

Laura Masiero, email: [e-mail]

References

Barachant, Alexandre, Stéphane Bonnet, Marco Congedo e Christian Jutten (2012). "Multiclass brain-computer interface classification by Riemannian geometry". In: IEEE Transactions on Biomedical Engineering. issn: 00189294. doi: 10.1109/TBME.2011.2172210.

Barachant, Alexandre e Marco Congedo (2014). “A Plug&Play P300 BCI Using Information Geometry”. url: http://arxiv.org/abs/1409.0107.

See Also

SuperTrial

LauraMasiero/RiemannR documentation built on Sept. 29, 2020, 9:51 p.m.