TRAMO/SEATS

The TRAMO (Time Series Regression with ARIMA Noise, Missing Observations, and Outliers) estimates and calculates models with missing data, ARIMA residuals, and with outliers of various types.

The SEATS model (Signal Extraction in ARIMA Time Series) decomposes time series into unobservable components based on ARIMA process.

These models are often used together as the alternative to other methods finding seasonality (such as X11 and X12).

Program implementation of both models is executed by Victor Gomez and Augustin Maravall for the Bank of Spain. The use of methods calculating in accordance with the TRAMO and SEATS models is available in Foresight Analytics Platform for Windows. To work with methods in Linux, an additional integration is required.