Seasonal Adjustment with TRAMO-SEATS in 'JDemetra+' 3.x

Interface to 'JDemetra+' 3.x (< https://github.com/jdemetra>) time series analysis software. It offers full access to options and outputs of 'TRAMO-SEATS' (Time series Regression with ARIMA noise, Missing values and Outliers - Signal Extraction in ARIMA Time Series), including 'TRAMO' modelling (ARIMA model with outlier detection and trading days adjustment). ARIMA = AutoRegressive Integrated Moving Average.

{rjd3tramoseats}

{rjd3tramoseats} offers full access to options and outputs of TRAMO-SEATS (rjd3tramoseats::tramoseats()), including TRAMO modelling (rjd3tramoseats::tramo()) and SEATS decomposition (rjd3tramoseats::seats_decompose()).

A specification can be created with rjd3tramoseats::tramo_spec() or rjd3tramoseats::tramoseats_spec() and modified with the following functions:

for pre-processing: rjd3toolkit::set_arima(), rjd3toolkit::set_automodel(), rjd3toolkit::set_basic(), rjd3toolkit::set_easter(), rjd3toolkit::set_estimate(), rjd3toolkit::set_outlier(), rjd3toolkit::set_tradingdays(), rjd3toolkit::set_transform(), rjd3toolkit::add_outlier(), rjd3toolkit::remove_outlier(), rjd3toolkit::add_ramp(), rjd3toolkit::remove_ramp(), rjd3toolkit::add_usrdefvar();
for decomposition: rjd3x13::set_x11();
for benchmarking: rjd3toolkit::set_benchmarking().
for decomposition: rjd3tramoseats::set_seats();
for benchmarking: rjd3toolkit::set_benchmarking().

Installation

Running rjd3 packages requires Java 21 or higher. How to set up such a configuration in R is explained here

🎉 {rjd3tramoseats} is now available on CRAN! 🎉

To install it, just launch the following command line:

install.packages("rjd3tramoseats")

To get the current development version of {rjd3tramoseats} from GitHub with:

# install.packages("remotes")
remotes::install_github("rjdverse/rjd3tramoseats")

Usage

Seasonal Adjustment with Tramo-Seats

library("rjd3tramoseats")
y <- rjd3toolkit::ABS$X0.2.09.10.M
ts_model <- tramoseats(y)
summary(ts_model$result$preprocessing) # Summary of tramo model
#> Log-transformation: yes 
#> SARIMA model: (0,1,1) (0,1,1)
#> 
#> Coefficients
#>           Estimate Std. Error  T-stat Pr(>|t|)    
#> theta(1)  -0.82783    0.02571 -32.196  < 2e-16 ***
#> btheta(1) -0.42554    0.06388  -6.661 9.01e-11 ***
#> ---
#> Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#> 
#> Regression model:
#>                   Estimate Std. Error T-stat Pr(>|t|)    
#> mon             -0.0109446  0.0034805 -3.145 0.001788 ** 
#> tue              0.0048940  0.0035307  1.386 0.166481    
#> wed              0.0001761  0.0034970  0.050 0.959867    
#> thu              0.0132928  0.0035330  3.763 0.000193 ***
#> fri             -0.0024801  0.0035383 -0.701 0.483748    
#> sat              0.0153509  0.0035171  4.365 1.62e-05 ***
#> lp               0.0410667  0.0101178  4.059 5.94e-05 ***
#> easter           0.0503888  0.0072698  6.931 1.69e-11 ***
#> AO (2000-06-01)  0.1681662  0.0299743  5.610 3.78e-08 ***
#> AO (2000-07-01) -0.1972348  0.0298664 -6.604 1.28e-10 ***
#> ---
#> Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#> Number of observations: 425, Number of effective observations: 412, Number of parameters: 13
#> Loglikelihood: 781.358, Adjusted loglikelihood: -2086.269
#> Standard error of the regression (ML estimate): 0.03615788 
#> AIC: 4198.538, AICc: 4199.452, BIC: 4250.811
plot(ts_model) # Plot of the final decomposition

Package Maintenance and contributing

Any contribution is welcome and should be done through pull requests and/or issues. pull requests should include updated tests and updated documentation. If functionality is changed, docstrings should be added or updated.