SeasCen, A Python-Based Platform for Time Series Modeling and Seasonal Adjustment

Whereas X-13ARIMA-SEATS (X-13) is widely used around the world to seasonally adjust economic time series, its continued longevity is jeopardized by the ongoing difficulty of maintaining its FORTRAN codebase. The FORTRAN language is no longer the platform of choice for new statistical software develo...

Full description

Saved in:
Bibliographic Details
Main Authors: Sara Alaoui, William Bell, Dan Haim, Demetra Lytras, Anup Mathur, Kathleen McDonald-Johnson, Tucker McElroy, Lijing Sun
Format: Article
Language:English
Published: Taylor & Francis Group 2025-12-01
Series:Data Science in Science
Subjects:
Online Access:https://www.tandfonline.com/doi/10.1080/26941899.2025.2531047
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Whereas X-13ARIMA-SEATS (X-13) is widely used around the world to seasonally adjust economic time series, its continued longevity is jeopardized by the ongoing difficulty of maintaining its FORTRAN codebase. The FORTRAN language is no longer the platform of choice for new statistical software development. Thus, the U.S. Census Bureau has been developing a new computer platform for seasonal adjustment that includes both a new Python-based user interface and an enhancement of X-13 based on wrapping the FORTRAN code in Python. This implementation also provides an analogous environment for RegComponent modeling capabilities, and facilitates use of X-13 by Linux users. RegComponent models have regression mean functions with residuals that follow ARIMA component time series models. These models can be used for seasonal adjustment, for signal extraction to improve repeated survey estimates, for modeling time-varying regression coefficients, and for other purposes. The new platform also includes additional capabilities for seasonal adjustment such as a tool to calculate X-11 symmetric filter weights for estimates of the various components: the nonseasonal (seasonally adjusted series), seasonal, seasonal-irregular, trend, and irregular.
ISSN:2694-1899