possibleFutures_benchmark: Possible future sample paths (multi-step) from residuals of a...

In smimodel: Sparse Multiple Index Models for Nonparametric Forecasting

Possible future sample paths (multi-step) from residuals of a fitted benchmark model

Description

Generates possible future sample paths (multi-step) using residuals of a fitted benchmark model through recursive forecasting.

Usage

possibleFutures_benchmark(
  object,
  newdata,
  bootstraps,
  exclude.trunc = NULL,
  recursive_colRange
)

Arguments

object	A fitted model object of the class backward, pprFit, or gaimFit.
newdata	The set of new data on for which the forecasts are required (i.e. test set; should be a tsibble).
bootstraps	Generated matrix of bootstrapped residual series.
exclude.trunc	The names of the predictor variables that should not be truncated for stable predictions as a character string.
recursive_colRange	The range of column numbers in newdata to be filled with forecasts.

Value

A list containing the following components:

firstFuture	A numeric vector of 1-step-ahead simulated futures.
future_cols	A list of multi-steps-ahead simulated futures, where each list element corresponds to each 1-step-ahead simulated future in firstFuture.

smimodel documentation built on April 8, 2026, 5:06 p.m.