model.matrix: GetPredictorMatrix

In BoomSpikeSlab: MCMC for Spike and Slab Regression

GetPredictorMatrix

Description

Extract the matrix of predictors.

Usage

  GetPredictorMatrix(object, newdata, na.action = na.omit, ...)

Arguments

object	An object of class glm.spike. The object must be a list with the following elements beta: a matrix of MCMC draws, with rows representing draws, and columns representing coefficients. xlevels: the levels of any contrasts present in the original training data. contrasts: the "contrasts" attribute of the original design matrix used to train the model. terms: the terms of the formula used to fit the original model.
newdata	A data frame, matrix, or vector containing the predictors needed to make a prediction. If newdata is a matrix it must have the same number of columns as length(object$beta), unless it is off by one and the model contains an intercept, in which case an intercept term will be added. If length(object$beta) == 1 (or 2, with one element containing an intercept) then newdata can be a numeric vector.
na.action	A function specifying what to do with NA's.
...	Extra arguments passed to model.matrix, in the event that newdata is a data frame.

Value

A matrix of predictor variables suitable for multiplication by object$beta.

Author(s)

Steven L. Scott

See Also

lm.spike SpikeSlabPrior plot.lm.spike predict.lm.spike

BoomSpikeSlab documentation built on May 29, 2024, 5:07 a.m.