LEiDA: Leading Eigenvector Dynamics Analysis
In maotai: Tools for Matrix Algebra, Optimization and Inference

LEiDA

R Documentation

Leading Eigenvector Dynamics Analysis

Compute the leading eigenvector dynamics analysis (LEiDA) of a multivariate time series as appearing in computational neuroscience.

LEiDA(X, TR, bp = c(0.01, 0.1), b_ord = 2)

`X`	A `(T,N)` matrix of multivariate time series data, where `T` is the number of time points and `N` is the number of ROIs.
`TR`	Repetition time (in secounds)
`bp`	Bandpass filter, a vector of length 2 with the lower and upper bounds of the bandpass filter in Hz. Default is `c(0.01, 0.10)`.
`b_ord`	Butterworth order, a positive integer. Default is 2.

a list containing

V: A (T,N) matrix of the leading eigenvector time series.
FCD_cos: A (T,T) matrix of the functional connectivity dynamics (FCD) using cosine similarity.
FCD_cor: A (T,T) matrix of the functional connectivity dynamics (FCD) using Pearson correlation.