BridgeChangeSim: Bridge Regression simulation

In jongheepark/BridgeChange: Bayesian Change-point models with Bridge prior for analyzing time-series and panel data.

Bridge Regression simulation

Description

Simulation code for univariate response change-point model with Bridge prior.

Usage

BridgeChangeSim(
  ntime = 500,
  predictor = 100,
  rho = 0,
  time.series = FALSE,
  standardize = TRUE,
  sign.change.tune = 2,
  sigma1 = 1,
  sigma2 = 2,
  train.ratio = 0.5,
  fitted.mse = TRUE,
  constant.p = 0.1,
  varying.p = 0.2,
  break.point = 0.5,
  positive.jump = FALSE,
  n.break = 1,
  intercept = FALSE,
  positive.jump.tune = 1,
  mcmc = 100,
  burn = 100,
  verbose = 100,
  thin = 1,
  N = 1,
  known.alpha = FALSE,
  dgp.only = FALSE
)

Arguments

ntime	Length of time series
predictor	Number of predictor
rho	correlation parameter 0 = no correlation.
time.series	TRUE if dgp is generated from autocorrelated series. rho is used as an autocorrelation coefficient.
sign.change.tune	tuning parameter for the size of parameter sign changes
sigma1	sigma 1
sigma2	sigma 2
train.ratio	The proportion of training data. (0, 1).
fitted.mse	If TRUE and n.break == 0, compute the MSE of fitted values against true responses. If FALSE, do the cross-validation test using training data.
constant.p	Proportion of constant parameters
varying.p	Proportion of time-varying parameters
break.point	Timing of a break between 0 and 1.
mcmc	=100
burn	=100
verbose	=100 Verbose
thin	Thinning
N	Number of cross-sectional units. If N=1, univariate time series data.
dgp.only	If TRUE, only data are generated and estimation steps are skipped.
corr.tune	tuning parameter for sx

Value

output

jongheepark/BridgeChange documentation built on Jan. 12, 2023, 4:48 p.m.