Defects fixed and structural changes to source code:
- Ensure AIC testing of multiple groups of user-defined regressors was handled correctly.
- Change error message for character string routine.
- Add check to input routines to test if names input exceed maximum length.
- Correct error message in automatic outlier input routine.
- Change array length for local variables in SEATS routines.
- Add variable for number of stock length of month regressors in regARIMA model.
- Add xtdtyp.cmn to source code, contains common blocks for irregular regression.
- Remove variables defined in xtdtyp.cmn from function and subroutine calls.
- Revise dimensions used for variables in spectral peak routines of SEATS code.
- Ensure updated variables are used in call to gtmdfl routine (reads regARIMA model coefficients from file).
- Update saving model file with scientific notation used to save user-defined regressors.
- Ensure enough forecasts are used to produce correct component values and forecasts for SEATS.
- Ensure that if maxorder is set to 0 for the seasonal component, the automatic model will have no seasonal MA terms.
- Add another pointer dictionary for additional tables to the getsav and getprt routines.
- Revise if - then - else structure in ansub1.f to ensure correct forecast parameters are used in the sigex routine of SEATS.
- Ensure correct table routine used in SEATS code.
- Change UDG code for SEATS skewness diagnostics to avoid duplication.
- Revise how trading day AIC test results are saved for UDG output.
- Ensure SEATS adjustment not attempted when ARIMA model not chosen and Missing Values are in series.
- Print warning message when all user-defined holiday regressors are deleted when chi-square testing is requested.
- Make correction for average duration of run statistic used in Q statistic.
For the vast majority of series, this version of X-13ARIMA-SEATS will produce X-11 seasonal adjustments that are identical to those generated by Version 1.1 Build 19 of X-13ARIMA-SEATS released earlier.
We are interested in your feedback. Please email your questions and comments to email@example.com.