Introduction

The primary focus of this module is on recoding data using the SDMX standard and Fusion Metadata Registry (FMR).

This module builds on the knowledge and skills previously acquired regarding SDMX structural modelling and structure management using FMR.

Throughout their lifecycle, data are subjected to many operations and transformations as they are massaged, aggregated, summarised, analysed, and customised for a wide range of outputs. Data are often collected in a highly granular level and disseminated at an aggregate level to facilitate key messaging and understanding. Data are repackaged according to the requirements of various international actors and projects, including OECD, IMF, ECB, Eurostat, World Bank, BIS, SDGs, UN Agencies, and many more. It is common that the same data needs to be described (structural model) in different ways for different processes and actors.

Additionally, organisations can rarely change all systems and databases to be SDMX-compliant with coding aligned to common, shared SDMX codelists over a short timeframe – even if they wanted to. Different sectors often use different coding approaches in their production systems. It is often necessary to take a phased approach and have a bridging strategy to facilitate implementing SDMX in some process areas and for some sectors while continuing business-as-usual elsewhere in the statistical system.

Wouldn’t it be great to be able to:

  • Adopt SDMX XLSX templates and the benefits that they offer without having to change all production processes and methods at the same time?
  • Easily transform data according to the various requirements of data partners such as SDGs and other reporting agencies in a straightforward and automated manner?
  • Recode data modelled according to one coding system (e.g. National Education System, Internal country coding) to another standard (e.g. ISCED, ISO3, M49) for collection, reporting, dissemination, or data production purposes?

There are capabilities in the SDMX standard which supports these use cases and facilitates the implementation of SDMX and remodelling data according to SDMX good practices in some processes/domains while in parallel, maintaining existing data structures, data formats, and coding in other parts of the statistical system.

In addition, these recoding capabilities may be bidirectional whereby the source to target mapping for scenario A may be reversed and become the target to source mapping in scenario B which adds even more value when using these capabilities in a statistical system and throughout various stages of the data lifecycle.

SDMX provides the ability to recode data according to various structural model mappings in an automated way. The SDMX method for doing so is by using structure maps and representation maps along with an SDMX transformation application. It is also possible to automate these transformations as a part of production workflows using the FMR data transformation web service.

What you’ll learn

Learning Objectives

In this module, you will:

  • Discover how to recode data using SDMX.
  • Learn the features and capabilities of FMR for recoding data.
  • Discover the FMR web services relating to recoding data.
  • Learn how to recode data using FMR.
  • Learn how to use FMR data transformation web services for automating recoding of data.

Prerequisites

You should already have a basic understanding of SDMX fundamentals and be familiar with the FMR user interface.

Units in this module

Learn structural modelling using statistical concepts and terms.