CONTENTS OF THIS FOLDER ——————————————

• HPCA_tutorial.R : A step-by-step implementation of HPCA_decomposition.R and the associated
  procedures described in "Hybrid Principal Components Analysis For Region-Referenced Longitudinal Functional EEG Data" by Scheffler et al. (2017).

• HPCA_decomposition.R : Function for performing HPCA decomposition including estimation of fixed effects, marginal covariance functions, marginal eigencomponents, subject-specific scores, variance components, and measurement error variance.

• HPCA_simulation.R : Function for simulating data for the HPCA_decomp function and the associated procedures referred to above.

• HPCA_bootstrap.R : Function for performing group-level inference for a given scalp region. The null hypothesis assumes that groups share a common region shift for a given region.

INTRODUCTION ——————————————

The contents of this folder allow for implementation of the HPCA decomposition described in "Hybrid Principal Components Analysis For Region-Referenced Longitudinal Functional EEG Data" by Scheffler et al. (2017). Users can simulate a sample data frame (HPCA_simulation.R) and apply the proposed HPCA decomposition (HPCA_decomposition.R). Further, we include tools to perform group-level inference via a bootstrap procedure (HCPA_bootstrap.R), allowing users to test whether the longitudinal functional stochastic process in a fixed region varies among groups. Detailed instructions on how to perform the aforementioned procedures, visualize results, and check the assumption of weak separability via a likelihood-ratio test on the correlation structure of the random effects are included in HPCA_tutorial.R.

REQUIREMENTS ——————————————

The included R programs require R 3.3.2 (R Core Team, 2016) and the packages listed in HPCA_tutorial.R.

INSTALLATION ——————————————

Load the R program files into the global environment and install required packages using commands in HPCA_tutorial.R