11 Transforming between versions of SDMX

Last modified by Helena on 2025/07/20 12:46

Contents

- 11.1 Scope
- 11.2 Compatibility and new DSD features

11.1 Scope

The scope of this section is to define both best practices and mandatory behaviour for specific aspects of transformation between different versions of SDMX.

11.2 Compatibility and new DSD features

The following table provides an overview of the backwards compatibility between SDMX 3.0 and 2.1.

SDMX 3.0 feature	SDMX 2.1 compatibility	Comments
Multiple Measures	Create a Measure Dimension Or Model Measures as Attributes	For a Measure Dimensions, all Concepts must reside in the same Concept Scheme
Arrays for values	Cannot be supported	Arrays are always reported in a verbose format, even if one value is reported
Measure Relationship	Can be ignored, as it does not affect dataset format
Metadata Attributes	Can be created as Data Attributes	Not for extended facets
Multilingual Components	Cannot be supported
No Measure	Can only be emulated by ignoring the Primary Measure value
Use extended Codelist	A new Codelist with all Codes must be created
Sentinel values	Cannot be supported in the DSD	Rules may be supported outside the DSD, in bilateral agreements

The following table illustrates forward compatibility issues from SDMX 2.1 to 3.0.

SDMX 2.1 feature	SDMX 3.0 compatibility	Comments
Measure Dimension	Create a Dimension with role ‘MEASURE’ Or Create multiple Measures from the Measure Dimension Concept Scheme	If the dataset has to resemble that of SDMX 2.1 Structure Specific, then the first option must be used
Primary Measure	Create one Measure with role ‘PRIMARY’; use id=”OBS_VALUE”