evalTerm {spikeSlabGAM}

Get summaries of the posterior (predictive) distribution of the linear predictor of a model term

Description

Get summaries of the posterior (predictive) distribution of the linear predictor of a model term

Usage

evalTerm(
  label,
  model,
  newdata = NULL,
  aggregate = mean,
  quantiles = c(0.1, 0.9),
  returnData = TRUE
)

Arguments

label

(character) the label of one of the terms in model.

model

a spikeSlabGAM object

newdata

data.frame on which to evaluate label. Defaults to NULL, in which case the term is evaluated on the original data.

aggregate

(function) a summary statistic that is applied over the mcmc-samples of the linear predictor. Defaults to mean.

quantiles

(numeric) a vector of quantiles for borders of credible regions of the linear predictor. Defaults to 10 and 90 percent quantiles, i.e. a (point-wise) 80 percent credible region.

returnData

should the relevant original variables be included in the returned data.frame? Defaults to TRUE.

Value

A data.frame that contains the relevant variables from newdata (if returnData is TRUE), the aggregate-summary and the requested quantiles.

Author(s)

Fabian Scheipl