create.data: Create Time Delay Embedding

In nunompmoniz/rewind: REsampling With Implicit DepeNDencies

Description

Time series forecasting models usually assume the existence of a degree of correlation between successive values of the series. A form of modelling such correlation is to use previous values of the series as predictors of future value(s), in a procedure known as time delay embedding. This process allows the us of standard regression tools on time series forecasting tasks.

Usage

create.data(ts, embed = 10)

Arguments

ts	the data structure with the time series data
embed	the size of the embed, defaults to 10

Value

a data frame with the time delay embedding of the time series

nunompmoniz/rewind documentation built on July 8, 2021, 12:25 a.m.