seqBART: Sequential Bayesian Additive Regression Trees Model

In sbart: Sequential BART for Imputation of Missing Covariates

Description

A flexible Bayesian nonparametric model that is used as imputation tool for missing covariates.

Usage

seqBART(xx, yy, datatype, type = 1, numskip = 199, burn = 1000, m = 200,
  sigdf = 3, sigquant = 0.9, kfac = 2)

Arguments

xx	Dataset of covariate matrix with missing values (NAs).
yy	Response (fully observed).
datatype	a vector indicating the type of covariates (0=continuous, 1=binary).
type	0=no reponse, 1=continuous response (linear regression used for imputation) and 2=binary response (logistic regression used for imputation)
numskip	number of iterations skipped
burn	number of iterations for burn-in
m	m value
sigdf	sig df value
sigquant	sign quant values
kfac	kd fac value

Value

Imputed Dataset Values

sbart documentation built on May 1, 2019, 7:23 p.m.