Statistical Disclosure Control |
|||||||||||||||||||||||||||||||||
Contact:Peter-Paul de WolfStatistics Netherlands P.O. Box 24500 2490 HA The Hague The Netherlands Phone: +31 70 337 5060 Last update: 26 August 2021 |
τ-ARGUS home pageIntroductionτ-ARGUS is a software program designed to protect statistical tables. The CASC-project took a previous version of τ-ARGUS, as developed during the SDC-project, as a starting point. After a test-phase (coordinated by IStat) version 3 was the final release of the CASC-project. However we have continued our efforts, also thanks to the ESSNet-project and at the end of the ESSNet project version 3.4.1 was available.
Major new extensions are the reading of SPSS-system files,
inclusion of the audit-routine, negative values in Modular and the
generation of a-priory files. Partly funded by Eurostat, a project was started on 20 December 2012 to port the at that time most recent version of τ-ARGUS to an Open Source version. The resulting version should contain the possibility to be run on a Windows platform as well as on a Linux/Unix platform. Moreover, the cell suppression, CTA and rounding methods should be able to make use of Open Solvers. τ-ARGUS software releases
See the User Manual of version 4.1 for more information on the use of τ-ARGUS. Since version 4.1.4 we provide two sets of binaries: one with and one without a bundled JRE7 distribution. To run the 32 bits Windows binary, Java 7 or higher in 32 bit version is needed (also on 64 bit machines). With the bundled JRE7 version you don't need to bother with installing a 32 bit version of Java on your (possibly 64 bit) machine: the ARGUS binary will automatically use the bundled runtime engine. As of version 4.2.0 we will only provide binaries with JRE bundled (Zulu Open JRE). You can download the latest Windows 32 bit binaries, build on Windows 10:
Older versions (Windows 32 bit):
You can download Linux binaries, Build on 32 bit Ubuntu 14.4.0:
You can download the sources of version 4.1.0 here, or
go to sdcTools on GitHub for most recent sources.
You can also still download the (Windows only) version previous to the Open Source version
here.
(UnZip the file in a temp-directory and run the install.bat procedure).
See the Manual of version 3.5 for more information. The following programs were created for the 3.5 version. They are not tested on later versions (4.1 and later).
For users of the SBS-format a conversion program to convert SBS-files to τ-ARGUS files
can be downloaded here
For users of SAS an interface has been written by the Swedisch colleagues as a deliverable of the ESSNet on SDC harmonisation. Optimisation solvers
For the calculation of an optimal pattern of secondary suppressions τ-ARGUS uses a LP-solver.
Due to the complexity of the mathematical problems involved we use high quality
commercial LP-solvers. Traditionally we
use the Xpress-MP solver made by FICO. Contact xpresssales@fico.com and reference your need for a τ-ARGUS specific license. Since we had to change the licencing mechanism of Xpress when going to Open Source, existing users should contact support@fico.com to get their licences renewed in compliance with the new situation. Alternatively CPlex can be used, e.g. for those users who already have a licence for CPlex and who want to use that solver. Although τ-ARGUS is free software (a result of a European 5th Framework project), these solvers are commercial products. You can download the τ-ARGUS software from this page, but you have to obtain a licence for one of these solvers yourself, directly from the suppliers. Since version 4.1.2 τ-ARGUS can also make use of Open Solvers. However, note that these solvers are not always able to deal with larger tables. Questions related to τ-ARGUS should be directed to Peter-Paul de Wolf. For the conditions to obtain Xpress-MP (including updates) see the 2010 pricelist here A summary of the installation procedure for τ-ARGUS can be found here For floating XPress licences a new licene manager has been develloped by FICO An upgrade for the XPress drivers for Windows XP can be found here For the conditions to obtain CPlex see here
|