reUseBenchmark {disaggR}R Documentation

Using an estimated benchmark model on another time-serie

Description

This function reapplies the coefficients and parameters of a benchmark on new time-serie.

Usage

reUseBenchmark(hfserie,benchmark,reeval.smoothed.part=FALSE)

Arguments

hfserie

the bended time-serie. If it is a matrix time-serie, it has to have the same column names than the hfserie used for the benchmark.

benchmark

a twoStepsBenchmark object, from which the parameters and coefficients are taken.

reeval.smoothed.part

a boolean of length 1. If TRUE, the smoothed part is reevaluated, hence the aggregated benchmarked serie is equal to the low-frequency serie.

Details

reUseBenchmark is primarily meant to be used on a serie that is derived from the previous one, after some modifications that would bias the estimation otherwise. Working-day adjustment is a good example. Hence, by default, the smoothed part of the first model isn't reevaluated ; the aggregated benchmarked serie isn't equal to the low-frequency serie.

Value

reUseBenchmark returns an object of class twoStepsBenchmark.

Examples

benchmark <- twoStepsBenchmark(turnover,construction) 
turnover_modif <- turnover
turnover_modif[2] <- turnover[2]+2
benchmark2 <- reUseBenchmark(turnover_modif,benchmark)

[Package disaggR version 1.0.1 Index]