R/urca
Navigate
Description Documentation Download Rcmdr add-in home	Description The functions contained in the package urca facilitate the application of unit root testing as well as the conduct of co-integration analysis. "urca" is considered as a core package for the task views "econometrics" and "finance". The package is purely written in R and S4-classes are employed. The current version is: 1.2-3 as of 28 September 2009. A ChangeLog file is shipped in its root directory such that any modifications are as transparent as possible to the user. You can view and download the changelog file by clicking on the following link: ChangeLog. The package has been written and is maintained by Bernhard Pfaff. Please adress correspondence to the email adress listed in the DESCRIPTION file of the package. The following unit root test are currently implemented: Augmented Dickey-Fuller Elliott-Rothenberg-Stock Phillips-Perron Schmidt-Phillips Kwiatkowski et al. Zivot-Andrews Co-integration analysis can be conducted by employing the methods proposed by: Phillips-Ouliaris Johansen Lütkepohl-Saikkonen-Trenkler Incidentally, the various tests for testing restrictions of the co-integrating and loading matrices in co-integrated vector autoregressive models have been implemented too. For more information and references see the package's documentation. Where applicable show-, summary- and plot-methods have been written for each of the objects generated by applying one of the functions. The package is shipped with numerous data sets, mostly the ones that have been utilised in the seminal papers such that the results therein can be replicated. Furthermore, a graphical user interface as an add-in to the Rcmdr is shipped in the /inst subdirectory of the package. To top Documentation The current documentation can be downloaded as pdf-file here. To top Download The latest tarball can be downloaded here: urca_1.2-3.tar.gz The latest windows binary can be downloaded here: urca_1.2-3.zip Aside of these sources, the package is available on CRAN and it is hosted on R-Forge: AICTS I (daily snapshots, forum, email list, etc.). To top Rcmdr add-in The /inst subdirectory of the package contains three files: Rcmdr-urca.R, Rcmdr-menus.txt and README. The latter explains how to set up the urca add-in into the Rcmdr. Once you have copied the file Rcmdr-urca.R into the subdirectory /etc of the Rcmdr package and inserted the menue-file you should see the following screen after launching the Rcmdr: To the left of the pull-down menu Help appears the urca menu. By clicking on it, you will find two sub-categories: unit root tests and co-integration analysis. Beneath each sub-menu appears a list of available tests and methods once you have selected either one. A data set must be loaded first before you can apply any of the functions. Below are some more screenshots of the Augmented Dickey-Fuller, Elliott-Rothenberg-Stock- and Zivot-Andrews test as well as the function's screen of estimating a vector error correction model. To top

R/urca

Navigate

Description
Documentation
Download
Rcmdr add-in

home

Description

The functions contained in the package urca facilitate the application of unit root testing as well as the conduct of co-integration analysis. "urca" is considered as a core package for the task views "econometrics" and "finance". The package is purely written in R and S4-classes are employed. The current version is: 1.2-3 as of 28 September 2009. A ChangeLog file is shipped in its root directory such that any modifications are as transparent as possible to the user. You can view and download the changelog file by clicking on the following link: ChangeLog. The package has been written and is maintained by Bernhard Pfaff. Please adress correspondence to the email adress listed in the DESCRIPTION file of the package.

The following unit root test are currently implemented:

Augmented Dickey-Fuller
Elliott-Rothenberg-Stock
Phillips-Perron
Schmidt-Phillips
Kwiatkowski et al.
Zivot-Andrews

Co-integration analysis can be conducted by employing the methods proposed by:

Phillips-Ouliaris
Johansen
Lütkepohl-Saikkonen-Trenkler

Incidentally, the various tests for testing restrictions of the co-integrating and loading matrices in co-integrated vector autoregressive models have been implemented too. For more information and references see the package's documentation.
Where applicable show-, summary- and plot-methods have been written for each of the objects generated by applying one of the functions. The package is shipped with numerous data sets, mostly the ones that have been utilised in the seminal papers such that the results therein can be replicated. Furthermore, a graphical user interface as an add-in to the Rcmdr is shipped in the /inst subdirectory of the package.

To top

Documentation

The current documentation can be downloaded as pdf-file here.

To top

Download

The latest tarball can be downloaded here: urca_1.2-3.tar.gz
The latest windows binary can be downloaded here: urca_1.2-3.zip
Aside of these sources, the package is available on CRAN and it is hosted on R-Forge: AICTS I (daily snapshots, forum, email list, etc.).

To top

Rcmdr add-in

The /inst subdirectory of the package contains three files: Rcmdr-urca.R, Rcmdr-menus.txt and README. The latter explains how to set up the urca add-in into the Rcmdr.
Once you have copied the file Rcmdr-urca.R into the subdirectory /etc of the Rcmdr package and inserted the menue-file you should see the following screen after launching the Rcmdr:

start screen of Rcmdr

To the left of the pull-down menu Help appears the urca menu. By clicking on it, you will find two sub-categories: unit root tests and co-integration analysis. Beneath each sub-menu appears a list of available tests and methods once you have selected either one.
A data set must be loaded first before you can apply any of the functions.
Below are some more screenshots of the Augmented Dickey-Fuller, Elliott-Rothenberg-Stock- and Zivot-Andrews test as well as the function's screen of estimating a vector error correction model.

Augmented Dickey-Fuller Elliott-Rothenberg-Stock Zivot-Andrews VECM

To top