BFDA Matlab Toolbox

Matlab toolbox for Bayesian functional data analysis using Gaussian-Wishart processes

Developed for Matlab 2014 or later

Required existing MATLAB packages for using this toolbox (included)

• bspline/ Matlab toolbox for implementing B-splines, can be downloaded from http://www.mathworks.com/matlabcentral/fileexchange/27374-b-splines.

• fdaM/ Matlab toolbox for standard functional data analysis, can be downloaded from http://www.psych.mcgill.ca/misc/fda/downloads/FDAfuns/Matlab/.

• mcmcdiag/ Matlab toolbox for diagonizing MCMC convergence, can be downloaded from http://becs.aalto.fi/en/research/bayes/mcmcdiag/.

• PACErelease2.11/ Matlab toolbox for PACE, written by Yao et. al, can be downloaded from http://www.stat.ucdavis.edu/PACE/download.html.

Add the path of required MATLAB packages

For example, run the following script lines (replace pwd with the directory of your MATLAB packages)

	addpath(genpath(cat(2, pwd, '/BFDA_SourceCode')))
	addpath(genpath(cat(2, pwd, '/bspline')))
	addpath(genpath(cat(2, pwd, '/fdaM'))) 
	addpath(genpath(cat(2, pwd, '/mcmcdiag'))) 
	addpath(genpath(cat(2, pwd, '/PACErelease2.11')))  

Main function of BFDA

• BFDA_SourceCode/BFDA.m -- Main function for smoothing and mean-covariance estimation.

• Examples/BayesianSmooth_eg.m Example script calling BFDA().

• Examples/FdRegress_eg.m Example script calling fdaM regression functions with output from BFDA().

Simulation related functions

• BFDA_SourceCode/sim_gfd.m Function to generate Gaussian functional data with common/partial grids, stationary/nonstationary covariance.

• BFDA_SourceCode/sim_gfd_rgrid.m Function to generate Gaussian functional data with random grids, stationary/nonstationary covariance.

• BFDA_SourceCode/sim_ngfd_rgrid.m Function to generate Non-Gaussian functional data with random grids, stationary/nonstationary covariance.