predict_betas

Given an object returned by <code>VCBART_ind</code> or <code>VCBART_cs</code> and matrices of continuous and categorical modifiers, returns MCMC samples of the coefficient functions evaluated the provided points.

Fits linear varying coefficient (VC) models, which assert a linear relationship between an outcome and several covariates but allow that relationship (i.e., the coefficients or slopes in the linear regression) to change as functions of additional variables known as effect modifiers, by approximating the coefficient functions with Bayesian Additive Regression Trees. Implements a Metropolis-within-Gibbs sampler to simulate draws from the posterior over coefficient function evaluations. VC models with independent observations or repeated observations can be fit. For more details see Deshpande et al. (2024) <doi:10.1214/24-BA1470>.

Sameer K. Deshpande

VCBART

Fit Varying Coefficient Models with Bayesian Additive Regression
Trees

Ray Bai

Cecilia Balocchi

Jennifer Starling

Jordan Weiss

predict_betas function

<dl> <dt>fit</dt>
<dd>A list returned by <code>VCBART_ind</code> or <code>VCBART_cs</code></dd> <dt>Z_cont</dt>
<dd>Matrix of continuous modifiers at which you wish to evaluate the covariate effect functions. Default is a 1x1 matrix, which signals that no continuous modifiers are required for these evaluations.</dd> <dt>Z_cat</dt>
<dd>Integer matrix of categorical modifiers at which you wish to evaluate the covariate effect functions. Default is a 1x1 matrix, which signals that no continuous modifiers are required for these evaluations.</dd> <dt>verbose</dt>
<dd>Boolean indicating whether the code should print its progress (<code>TRUE</code>). Default is <code>TRUE</code>.</dd></dl>

Arguments

Compute posterior predictive evaluates of covariate effect functions. — predict_betas

<dl>

 <dt>fit</dt>
<dd>A list returned by <code>VCBART_ind</code> or <code>VCBART_cs</code></dd>

 <dt>Z_cont</dt>
<dd>Matrix of continuous modifiers at which you wish to evaluate the covariate effect functions. Default is a 1x1 matrix, which signals that no continuous modifiers are required for these evaluations.</dd>

 <dt>Z_cat</dt>
<dd>Integer matrix of categorical modifiers at which you wish to evaluate the covariate effect functions. Default is a 1x1 matrix, which signals that no continuous modifiers are required for these evaluations.</dd>

 <dt>verbose</dt>
<dd>Boolean indicating whether the code should print its progress (<code>TRUE</code>). Default is <code>TRUE</code>.</dd>

</dl>

predict_betas: Compute posterior predictive evaluates of covariate effect functions.

Description

Usage

Value

Arguments