RJDemetra 0.2.8.9000
All notable changes to this project will be documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
Unreleased
Changed
- Add test on the frequency of the input time series to avoid misleading error (#165).
0.2.8 - 2024-12-12
Changed
Improvement of warning message importing the model when no data in a SaItem (#61).
get_model,get_jmodelandget_jspec.sa_itemhave now an argumenttypeto define which specification to import (domain, estimation or point). By default the domain specification is extracted (as before).
Fixed
Correction of bug of
add_sa_item()of models created byjtramoseats()with external variables.Correction of bug of
get_jmodel()when model contains user-defined regressors (calendar or regressors) with fixed coefficients (#157).Correction of bug for annual data for
regarima()functions.
0.2.7 - 2024-10-09
Changed
URL to github repository updated (github.com/jdemetra replaced by github.com/rjdverse).
results of
user_defined_variables()updated..jars updated.
Added
- benchmarking option added to
x13_spec()andtramoseats_spec()and in output ofx13()andtramoseats().
0.2.6 - 2024-03-20
Added
possibility to export last msr for monthly data (issue #122).
possibility to export X-11 some components:
y_cmp,y_cmp_f,t_cmp,t_cmp_f,sa_cmp,s_cmp,s_cmp_f, andi_cmp.
Fixed
correction when importing models containing ramp regressors when the frequency is not 12 (monthly).
correction of
get_jmodel()with empty multiprocessings.
0.2.5 - 2024-02-04
0.2.4 - 2023-10-18
Added
- New function
get_all_names()andget_position().
Fixed
- correction of
save_workspace()andload_workspace()when using relative path.
0.2.3 - 2023-06-04
0.2.2 - 2023-04-27
0.2.1 - 2022-11-15
0.2.0 - 2022-11-15
Changed
Java 17 compatibility.
Change of javanamespace to avoid name clashes with jd+ 3.0.
Default parameters of the span specification
d0andd1as NA for clarity (issue #102).
Fixed
Bug with x11 introduced in RJDemetra 0.1.9 (issue #99).
easter.enablednot working since RJDemetra 0.1.9.
0.1.9 - 2021-11-17
0.1.8 - 2021-11-17
Fixed
- Java version restriction: JDemetra+ and RJDemetra are not compatible with Java 16 and higher. The compatibility with those versions of Java will be possible from next release of JDemetra+.
Added
-
java_ncoreoption added to limit the number of cores used in Java to two to be sure to respect CRAN policies (to remove the option, useoptions(java_ncore = NULL)). However, it should not be necessary since RJDemetra shouldn’t use multithread (issue #89).
0.1.7 - 2021-06-07
Fixed
fixed coefficients with user-defined calendar regressors can now correctly be used (issue #87).
there was a bug in
add_sa_item(more precisely incomplete_dictionary.SA) when a userdefined variable was already in the workspace but with a different suffix.
0.1.6 - 2020-08-11
Fixed
x11.seasonalmaargument wasn’t taken into account (issue #78).ipi_c_euupdated: the previous data were calendar adjusted data, they are now unadjusted (neither seasonally adjusted nor calendar adjusted data).the print result of the combined test was incorrect.
0.1.5 - 2020-03-12
0.1.4 - 2020-01-10
Fixed
- Error while loading a workspace with metadata (issue #53).
- The degree of freedom were wrong with the
summaryfunctions. -
jregarimafunction is now exported. - The message was not complete when there was an error importing a model with
get_model. - The trading-days specification was not correctly specified with TRAMO-SEATS.
Added
- Parameter
seats.predictionLengthadded totramoseats_spec. - Parameters
x11.calendarSigmaandx11.sigmaVectoradded tox13_spec. - Parameter
ylimadded toplotfunctions (issue #60). - New generic functions:
logLik,vcov,df.residual,nobs,residuals(linked to issue #56). - New pre-specified specification “X11” (no pre-processing; linked to issue #59).
- Ramp effects are now imported.
-
add_sa_itemnow compatible withjSAobject. - When a model is added to a workspace with
add_sa_item, the external regressors are renamed only if they don’t already exist in the workspace. - Warning added when
tradingdays.option = "UserDefined"andtradingdays.autoadjust,tradingdays.leapyearortradingdays.stocktdtradingdaysare defined in a new specification (because they are currently ignored). (issue #67)
0.1.3 - 2019-06-26
0.1.2 - 2019-04-18
Changed
- All _def functions are now deprecated and replaced by the functions with the same name but without _def. Use:
x13_specinstead ofx13_spec_def,x13instead ofx13_def,tramoseats_specinstead oftramoseats_spec_def,tramoseatsinstead oftramoseats_def,regarima_spec_tramoseatsinstead ofregarima_spec_def_tramoseats,regarima_tramoseatsinstead ofregarima_def_tramoseats,regarima_x13instead ofregarima_def_x13andregarima_spec_x13instead ofregarima_spec_def_x13. -
objectargument renamed byspecinx13_spec,tramoseats_spec,regarima_spec_x13andregarima_spec_tramoseats.
Added
- Parameter
preliminary.checkadded to the specifications functions (regarima_spec_tramoseats,tramoseats_spec,regarima_spec_x13andx13_spec). By default (preliminary.check = TRUE), JDemetra+ checks the quality of the input series and exclude highly problematic ones: e.g. these with a number of identical observations and/or missing values above pre-specified threshold values. Whenpreliminary.check = FALSE, the thresholds are ignored and process is performed, when possible. (issue #39) - Error message returned when the seasonal adjustment fails due to the preliminary.check.
- Possibility to use user-defined calendar regressors. To do it use
tradingdays.option = "UserDefinedand add new regressors variables (usrdef.varEnabled = TRUEto enable user-defined regressors andusrdef.varto define the regressors) usingusrdef.varType = "Calendar". -
usrdef.varTypeargument is recycled with the number of variables defined in theusrdef.varparameter. - News functions to only get the Java object from a seasonal adjustment or a pre-adjustment method:
jx13,jtramoseats,jregarima,jregarima_x13,jregarima_tramoseatsandget_jmodel. Therefore, there is no formatting and the computation is faster than the non ‘j’ functions (x13,tramoseats,regarima,regarima_x13,regarima_tramoseatsandget_model). To manipulate these objects, there are three functions:get_dictionaryto get the indicators that can be extracted,get_indicatorsto extract these indicators andjSA2Rto get the formatted R model.
Fixed
-
x11.fcastcan now be set to 0 or 1 (issue #42).