AHANM2024

AM_main: the main estimation file.
AM_01: models with 100 variables.
AM_02: models with correlated components.
AM_real01: the first real application.
AM_real02: the second real application.
GM3: construct k by k G matrix.
selectionM: construct n by k selection D matrix.
igaurnd: inverse-Gaussian random number generator.
tnormrnd: truncated normal random number generator.
lpost_d: function that calculates the log density of parameter d.
Boston.txt: data file associated with AM_real01.
diabetes.data: data file associated with AM_real02.
AM_03: prediction comparison with linear regression and lasso.
AM_03_dense.mat: prediction results, dense model.
AM_03_sparse.mat: prediction results, sparse model.
AM_real03: the third real application.
CPILFESL.csv: data file associated with AM_real03, quarterly.
CPILFESL(1).csv: data file associated with AM_real03, monthly.
gam01: large GAM using cubic spline.
gam02: small GAM using cubic spline.
gam03: cubic spline prediction comparison with linear regression and lasso.
AM_03_dense_mgcv.RData: cubic spline prediction results, dense model.
AM_03_sparse_mgcv.RData: cubic spline prediction results, sparse model.
gam04: adding MARS to gam03.
AM_MARS: DGP is from equation 56 in 'MARS' paper by Friedman 1991.
AM_Threshold: DGP with threshold.
AM_main_hs: added Horseshoe prior.