DDI DExT

The DDI-DExT (Data Exchange Tools) is a collaborative project between the UK Data Archive and the Open Data Foundation for the development of data archiving and conversion tools. Supported by the UKDA Data Exchange Tools (DExT) initiative, the project aims at the development of open source utilities for the conversion of quantitative data into a standard archiving format and export to common statistical packages or database systems. The application is based on the DDI specification and supports metadata export to DDI 2.0 and DDI 3.0. This initial product is a proof of concept and focuses on SPSS as the input data format with data export capabilities to ASCII, SAS, Stata, and SPSS. The output of the DDI DExT project has been donated to the DDI Foundation Tools Program.

Download

The DDI-DExT application is available for download from the project repository on ODaF Forge. Two packages are made available: the main DDI-DExT application and the SPSS Reader.

The DDI-DExT main tool is a stand alone Eclipse RCP application that manages "archive" projects that hold a collection of data files. The data can be imported from most SPSS files and are preserved in fixed ASCI format accompanied by DDI 3.0 based metadata to capture the file and variable level information extracted from SPSS. Archived data files can be redistributed to users for various statistical packages (currently SAS, Stata and SPSS). The export wizard attempts as much as possible to take into account the differences or incompatibilities between the software packages (like variable name lengths, string and date formats, etc.). It also has the ability to simultaneously generate multiple output formats for multiple files and save the resulting files in either a folder or a compressed ZIP file. The output can also contain DDI 3.0 and DDI 2.1 XML metadata for the generated files.

The underlying SPSS Reader Java package is also made available for developers interested in exporting data and metadata out of SPSS files. The package has the ability to generate data in fixed, delimited and CSV ASCII formats as well as export metadata in DDI 3.0 and DDI 2.1 XML formats. The package has been tested with a wide range of SPSS files produced on different platforms and varying from version 4 to 16. While minor issues remain to be addressed, the library is capable of reading most SPSS data files.

Developers / Contributors

Source code for the DDI-DExT tools is available in project CVS on ODaF Forge

Known issues

  • Need to align on DDI 3.0 once specification is officially released end of May (current implementation based on candidate release)
  • Data loss in fixed ASCII if variable width specified in SPSS is not wide enough to hold highest/lowest value
  • Cannot import files produced on Solaris and with the SPSS Writer Java package available on SourceForge (trivial fix)
  • Slight offset when importing data in the dd-mmm-yyyy hh:mm:ss.ss format

Roadmap

  • Support for additional input formats such as Stata, SAS, SPSS Portable or Nesstar
  • Support for additional setup file formats such as R, Excel or SQL Databases
  • Data export to a “generic” ASCII format that uses ISO standard dates, general number formats.
  • Addition of summary statistics to the DDI metadata

License

All DDI-DExT products are made available under the GNU Lesser General Public License version 2.1 or later.

Contact

For questions, comments, ideas, suggestions or contributions related to DDI-DExT, please contact pheus@opendatafoundation.org.