fdlm_flc: Factor Loading Curve Sampling Algorithm

In drkowal/dfosr: Dynamic Function-on-Scalars Regression

Description

Sample the factor loading curve basis coefficients subject to an orthonormality constraint. Additional linear constraints may be included as well.

Usage

fdlm_flc(BtY, Beta, Psi, BtB, Omega, BtCon = NULL, lambda, sigmat2)

Arguments

BtY	J x T matrix B.t()*Y for basis matrix B
Beta	T x K matrix of factors
Psi	J x K matrix of previous factor loading curve coefficients
BtB	J x J matrix of B.t()*B
Omega	J x J prior precision/penalty matrix
BtCon	(optional) J x Jc matrix of additional constraints, pre-multiplied by B.t()
lambda	K-dimensional vector of prior precisions
sigmat2	T-dimensional vector of time-dependent observation error variances

Value

Psi J x K matrix of factor loading curve coefficients

Note

This is a wrapper for Rcpp functions for the special cases of K = 1 and whether or not additional (linear) constraints are included, i.e., whether or not BtCon is non-NULL.

drkowal/dfosr documentation built on May 7, 2020, 3:09 p.m.