Using R in Seasonal Adjustment Production of Official Statistics

• Activity Number: 196 - Time Series Methods with Seasonal, Monthly, and Daily Data
• Type: Contributed
• Date/Time: Tuesday, August 4, 2020 : 10:00 AM to 2:00 PM
• Sponsor: Business and Economic Statistics Section
• Author(s): Brian Monsell*
• Keywords: Diagnostics; high-frequency time series; signal extraction; regARIMA models; fractional ARIMA models; graphics

Abstract

This talk details my work at the Bureau of Labor Statistics in two areas – developing R packages for use in seasonal adjustment production, and evaluating several methods of weekly seasonal adjustment for weekly time series. The packages I’m developing extract seasonal adjustment diagnostics from the output of the seasonal package, duplicate the functionality of the TERROR utility produced by Agustin Maravall, and generate diagnostic plots used by analysts at BLS. In addition, I have examined the use of R routines developed at the Federal Reserve Board (implementing the Cleveland method) and the creators of JDemetra+ (implementing a canonical decomposition of a fractional airline model) and compared them to the current method at use at BLS.