jpetrovich02/BFOFR: Bayesian Function-on-Function Regression for Multilevel Data

Implements bayesian function-on-function regression for multilevel data as described in Meyer et al. (2015). The method involves first projecting the functions X and Y onto a wavelet basis and compressing the wavelet coefficients according to the method described in Morris et al. (2011). This is followed by (optionally) further reducing the dimension of X by performing PCA and keeping only those principal components required to explain a specified proportion of the variation. An MCMC sampler is then used to estimate the beta coefficients. This is performed using the WFMM executable file created by The University of Texas MD Anderson Cancer Center. Finally, the beta coefficients are projected back to the original data space for interpretation and inferential procedures.

install.packages("remotes")
remotes::install_github("jpetrovich02/BFOFR")