in_sample: Producing the in sample predictions of a prais-lm regression

In arnaud-feldmann/disaggR: Two-Steps Benchmarks for Time Series Disaggregation

Producing the in sample predictions of a prais-lm regression

Description

The function in_sample returns in-sample predictions from a praislm or a twoStepsBenchmark object.

Usage

in_sample(object, type = "changes")

Arguments

object	an object of class "praislm" or "twoStepsBenchmark".
type	"changes" or "levels". The results are either returned in changes or in levels.

Details

The functions plot and autoplot can be used on this object to produce graphics.

The predicted values are different from the fitted values :

• they are eventually reintegrated.

• they contain the autocorrelated part of the residuals.

Besides, changes are relative to the latest benchmark value, not the latest predicted value.

Value

a named matrix time series of two columns, one for the response and the other for the predicted value. A "tscomparison" class is added to the object.

See Also

in_disaggr in_revisions in_scatter plot.tscomparison

Examples

benchmark <- twoStepsBenchmark(turnover,construction,include.rho = TRUE)
plot(in_sample(benchmark))

arnaud-feldmann/disaggR documentation built on July 21, 2024, 8:15 a.m.