tsBlocks: Construct predictor blocks for time series models
Description
Construct blocks of original and lagged values for
autoregressive time series models with exogenous inputs.
The typical use case is to supply the output as
newdata argument to the
predict method of robust
groupwise least angle regression models.
Usage
tsBlocks(x, y, p = 2, subset = NULL, intercept = TRUE)
Arguments
x
a numeric matrix or data frame containing the
exogenous predictor series.
y
a numeric vector containing the response
series.
p
an integer giving the number of lags to include
(defaults to 2).
subset
a logical or integer vector defining a
subset of observations from which to construct the matrix
of predictor blocks.
intercept
a logical indicating whether a column of
ones should be added to the matrix of predictor blocks to
account for the intercept.
Value
A matrix containing blocks of original and lagged values
of the time series y and x.