Cadernos do IME: Série Estatística (Mar 2017)

USING R TO TEACH SEASONAL ADJUSTMENT

  • Pedro Costa Ferreira,
  • Daiane Marcolino de Mattos

DOI
https://doi.org/10.12957/cadest.2016.25077
Journal volume & issue
Vol. 40, no. 1
pp. 19 – 33

Abstract

Read online

This article shows, using R software, how to seasonally adjust a time series using the X13-ARIMA-SEATS program and the seasonal package developed by Christoph Sax. In addition to presenting step-by-step seasonal adjustment, the article also explores how to analyze the program output and how to forecast the original and seasonally adjusted time series. A case study was proposed using the Brazilian industrial production. It was verified that the effect of Carnival, Easter and working days improved the seasonal adjustment when treated by the model.

Keywords