generics: Generic Methods for 'bartcFit' Objects

In lurui0421/BartCause: Causal Inference using Bayesian Additive Regression Trees

Description

Visual exploratory data analysis and model fitting diagnostics for causal inference models fit using the bartc function.

Usage

## S3 method for class 'bartcFit'
fitted(object,
                          value = c("est", "y", "y0", "y1", "indiv.diff", "p.score", "p.weights"),
                          sample = c("inferential", "all"),
                           ...)

extract(object, ...)

## S3 method for class 'bartcFit'
extract(object,
                           value = c("est", "y", "y0", "y1", "indiv.diff", "p.score", "p.weights"),
                           sample = c("inferential", "all"),
                           combineChains = TRUE,
                           ...)

refit(object, newresp, ...)

## S3 method for class 'bartcFit'
refit(object,
                         newresp = NULL,
                         commonSup.rule = c("none", "sd", "chisq"),
                         commonSup.cut  = c(NA_real_, 1, 0.05),
                         ...)

predict(object, newdata, ...)

## S3 method for class 'bartcFit'
predict(object,
                           newdata,
                           value = c("y1", "y0", "indiv.diff", "p.score"),
                           combineChains = TRUE,
                           ...)

Arguments

object	Object of class bartcFit.
value	Which quantity to return. See details for a description of possible values.
sample	Return information for either the "inferential" (e.g. treated observations when the estimand is att) or "all" observations.
combineChains	If the models were fit with more than one chain, results retain the chain structure unless combineChains is TRUE.
newresp	Not presently used, but provided for compatibility with other definitions of the refit generic.
newdata	Data corresponding to the confounders in a bartc fit.
commonSup.rule, commonSup.cut	As in bartc
...	Additional parameters passed up the generic method chain.

Details

fitted returns the values that would serve as predictions for an object returned by the bartc function, while extract instead returns the full matrix or array of posterior samples. The possible options are:

• "est" - the estimate itself, e.g. ATE

• "y" - predictions under the observed treatment condition, i.e. \hat{y}_i(1) * z_i + \hat{y}_i(0) * (1 - z_i).

• "y0" - predictions for all observations under the control

• "y1" - predictions for all observations under the treatment

• "indiv.diff" - for all observations the individual treatment effect estimates, i.e. \hat{y}_i(1) \hat{y}_i(0).

• "p.score" - probability that each observation is assigned to the treatment group

• "p.weights" - weights assigned to each individual difference if the response method is "p.weight"

refit exists to allow the same regressions to be used to calculate estimates under different common support rules. To refit those models on a subset, see the examples in bartc.

predict allows the fitted model to be used to make predictions on an out-of-sample set. Requires model to be fit with keepTrees equal to TRUE.

Value

For fitted, extract, and predict, a matrix, array, or vector depending on the dimensions of the result and the number of chains. For the following, when n.chains is one the dimension is dropped.

• "est" - when fitted, a scalar; when extract, n.samples x n.chains

• "y", "y0", "y1", "indiv.diff", "p.weights" - when fitted, a vector of length equal to the number of observations (n.obs); when extract or predict, a matrix or array of dimensions n.obs x n.samples x n.chains.

• "p.score" - depending on the fitting method, samples may or not be present. When samples are absent, a vector is returned for both functions. When present, the same as "y".

For refit, an object of class bartcFit.

Author(s)

Vincent Dorie: [e-mail].

See Also

bartc

Examples

## fit a simple linear model

lurui0421/BartCause documentation built on July 4, 2019, 1:02 p.m.