Run the ‘JWSACruncher’ with rjwsacruncher

Launch the ’JWSACruncher”

To use rjwsacruncher you need to download the JWACruncher. It is available on https://github.com/jdemetra/jdemetra-app/releases (for JDemetra+ 2.x.y versions) or https://github.com/jdemetra/jdemetra-app/releases (for JDemetra+ 3.x.y versions). It can also be downloaded with the function download_cruncher():

library(rjwsacruncher)
# Directory where to save the JWSACruncher:
directory <- tempdir()
download_cruncher(directory)
# for JDemetra+ 3.x.y versions :
download_cruncher(directory, v3 = TRUE)

To configure the JWSACruncher with a portable version of Java you can use the function configure_jwsacruncher(). See ?configure_jwsacruncher for more informations.

To run the JWSACruncher you need three files:

  • a “.param” file containing the refresh policy and the items of the seasonal adjustment to export;
  • a valid JDemetra+ workspace;
  • the path to the JWSACruncher.

In rjwsacruncher, there are four functions associated to run the JWSACruncher:

  • create_param_file() or list2param_file() to create the “.param” parameter file;
  • cruncher() to run the JWSACruncher on a workspace from a parameter file;
  • cruncher_and_param() to run the ‘JWSACruncher’ on a workspace while creating the parameter file;
  • update_workspace() to update a workspace without exporting the results.

Create the parameter file with create_param_file()

The parameters of the function create_param_file() are those described in the wiki of the JWSACruncher: https://github.com/jdemetra/jwsacruncher/wiki. The three most important parameters of create_param_file() are:

  1. policy the refresh policy (see table below).
Refresh policies
Option on JDemetra+ Option for the JWSACruncher Description
Partial concurrent adjustment -> Fixed model current The ARIMA model, outliers and other regression parameters are not re-identified and the values of all parameters are fixed. The transformation type remains unchanged.
Partial concurrent adjustment -> Estimate regression coefficients fixedparameters (or fixed) The ARIMA model, outliers and other regression parameters are not re-identified. The coefficients of the ARIMA model are fixed, other coefficients are re-estimated. The transformation type remains unchanged.
Partial concurrent adjustment -> Estimate regression coefficients + Arima parameters parameters (by default) The ARIMA model, outliers and other regression parameters are not re-identified. All parameters of the RegARIMA model are re-estimated. The transformation type remains unchanged.
Partial concurrent adjustment -> Estimate regression coefficients + Last outliers lastoutliers The ARIMA model, outliers (except from the outliers in the last year of the sample) and other regression parameters are not re-identified. All parameters of the RegARIMA model are re-estimated. The outliers in the last year of the sample are re-identified. The transformation type remains unchanged.
Partial concurrent adjustment -> Estimate regression coefficients + all outliers outliers The ARIMA model and regression parameters, except from outliers) are not re-identified. All parameters of the RegARIMA model are re-estimated. All outliers are re-identified. The transformation type remains unchanged.
Partial concurrent adjustment -> Estimate regression coefficients + Arima model stochastic Re-identification of the ARIMA model, outliers and regression variables, except from the calendar variables. The transformation type remains unchanged.
Concurrent complete (or concurrent) Re-identification of the whole RegARIMA model.
  1. matrix_item character containing the items of the matrix output. By default, the items defined in the option default_matrix_item are exported. To change it we can change the option default_matrix_item or the parameter matrix_item:
# To get the default parameters
getOption("default_matrix_item")
# To change the default parameters to, for example, only export
# the information criteria:
options(default_matrix_item = c("likelihood.aic",
                                "likelihood.aicc",
                                "likelihood.bic",
                                "likelihood.bicc"))
  1. tsmatrix_series character containing the names of the times series to export. By default, the items defined in the option default_tsmatrix_series are exported. To change it we can change the option default_tsmatrix_series or the parameter tsmatrix_series:
# To get the default parameters
getOption("default_tsmatrix_series")
# To change the default parameters to, for example, only export
# the seasonally adjusted series and its forecasts:
options(default_tsmatrix_series = c("sa", "sa_f"))

For more informations, see the help of the function: ?create_param_file. Below some examples to create the parameter file:

export_dir <- tempdir()
# To create the file parameters.params in the directory export_dir with
# the refresh policy "lastoutliers" and the others default parameters:
create_param_file(dir_file_param = export_dir,
                  policy = "lastoutliers")

# If the options "default_matrix_item" and "default_tsmatrix_series" were
# changed to only export the information criteria, the seasonally adjusted series and its forecasts, the previous code is equivalent to:
create_param_file(dir_file_param = export_dir,
                  policy = "lastoutliers",
                  matrix_item = c("likelihood.aic", "likelihood.aicc",
                                  "likelihood.bic", "likelihood.bicc"),
                  tsmatrix_series = c("sa", "sa_f"))

Parameter files can be read with read_param_file() which returns a list than can also be modified and exported with list2param_file():

param_f <- read_param_file(file.path(export_dir, "parameters.param"))
str(param_f)
## List of 7
##  $ config         :List of 4
##   ..$ bundle       : chr "10000"
##   ..$ csv_layout   : chr "list"
##   ..$ csv_separator: chr ";"
##   ..$ ndecs        : chr "6"
##  $ policy         : chr "lastoutliers"
##  $ refreshall     : logi TRUE
##  $ output         : NULL
##  $ matrix_item    : chr [1:4] "likelihood.aic" "likelihood.aicc" "likelihood.bic" "likelihood.bicc"
##  $ tsmatrix_series: chr [1:2] "sa" "sa_f"
##  $ paths_path     : NULL

The default parameters files of JDemetra+ 2.x.y and JDemetra+ 3.x.y can be retrieved with default_param_file().

To run the JWSACruncher

To run the JWSACruncher with cruncher() or cruncher_and_param(), you have to specify the path to the JWSACruncher (cruncher_bin_directory parameter) and the path to the workspace (workspace parameter).

By default, the path to the JWSACruncher is the value of the option cruncher_bin_directory: therefore you only have to change this option once so that it applies to all the future running. The path to specify is the folder containing the jwsacruncher.bat file which is under the “Bin” folder of the JWSACruncher. Thus, if it is installed in D:\jdemetra-cli-2.2.2, the file jwsacruncher.bat will be under D:\jdemetra-cli-2.2.2\bin and you have to change the cruncher_bin_directory option as follows:

options(cruncher_bin_directory = "D:/jdemetra-cli-2.2.2/bin/")

If no workspace is specified, a dialog box opens to select it.

The cruncher_and_param() function allows to create a temporary parameter file with create_param_file() and then run the JWSACruncher with cruncher(). In addition to the parameters available in these two functions, cruncher_and_param() allows to rename the output folder containing the workspace results so that they are equal to the names of the multi-documents displayed in the JDemetra+ software with the parameter rename_multi_documents (equals to FALSE by default). Below are some examples:

# The following code updates the workspace "workspace", that is under the folder D:/, 
# with the refresh policy "lastoutliers". Others parameters are the default ones of create_param_file().
# In particular, the exported parameters are those of the options
# "default_matrix_item" options and "default_tsmatrix_series" and the results 
# will be under D:/workspace/Output/.
cruncher_and_param(workspace = "D:/workspace.xml",
                   policy = "lastoutliers")

# Use the parameter "outpout" to change the folder that will contains the results
cruncher_and_param(workspace = "D:/workspace.xml",
                   output = "D:/Results/",
                   policy = "lastoutliers")

# To change the names of the folders containing the outputs so that they are equal
# to the names of the multi-documents displayed in JDemetra+, use the parameter
# "rename_multi_documents = TRUE". The parameter "delete_existing_file = TRUE"
# allows to delete any existing folders with the same name as the multi-documents.
cruncher_and_param(workspace = "D:/workspace.xml",
                   rename_multi_documents = TRUE,
                   delete_existing_file = TRUE,
                   policy = "lastoutliers")

# To see the other parameters:
?cruncher_and_param