IMF SDMX Central

The file name of the currently loaded file

The file format of the loaded dataset.

The number of this Dataset from the file

A Data Structure [Definition] (DSD) defines the structure of a dataset in terms of the dimensionality and attributes, collectively known as Components. The DSD can also define the allowable reported values for each component, known as Representation.

Allowable content may be free text, restricted by type, length, text patterns, and more. Allowable content may also be defined as an enumeration of allowable values, which will be defined in a separate structure known as a Codelist.

A Dataflow can restrict the allowable content for reported data beyond that defined by the DSD. These additional restrictions (called Constraints in SDMX) will be taken into account if the Dataflow is selected.

A Provision Agreement can restrict the allowable content for reported data beyond that defined by the DSD. These additional restrictions (called Constraints in SDMX) will be taken into account if the Provision Agreement is selected.

A Data Provider can have the allowable content restricted for reported data beyond that defined by the DSD, or Dataflow. These additional restrictions (called Constraints in SDMX) will be taken into account if the Data Provider is selected.

The number of series in the file

The number of observations in the file

The number of groups in the file. Note that SDMX Groups, if used, attach Attributes at a higher level than the Series i.e., they are expressed as a sub-set of the dimensionality of the series and are therefore a partial key and any Attributes therefore relate to multiple series.

For some data formats (e.g. SDMX-ML or SDMX-EDI) the header section contains the sender id. This field can be used to set the value the sender id will take.

If a Structure Map has been defined to map the dataset's Data Structure or Dataflow to a different Data Structure, then this drop down will be populated with Data Structures / Dataflows that can be mapped. On export the dataset will be converted in terms of its structure, and representation e.g. the dataset can be converted to conform to a different Data Structure Definition, and into a different data format

Requesting a mapping report forces the output to be a Zip file. This file will contain 3 files:

• out - The mapped output in the format of your choice.
• unmapped - The unmapped output in the same format as the mapped file.
• metrics.json - A JSON report stating the number of Series and Observations in both the mapped and unmapped files as well as a breakdown of why the elements were unmapped.

If a row of data is mapped successfully it can be subject to Data Validation. This will ensure that certain business rules are applied to the mapped data.

For example if the target structure has a FREQ dimension, the TIME_PERIOD must conform to that value. So a 4 digit-year must have a FREQ value of 'A'.

If you do not want validation to be perfomed, please ensure this setting is set so that "Invalid Data is Included in the Mapped Dataset".

Ensures the data format is consistent with the specification. These errors are specific to the data format loaded (XML/Excel/CSV). IMF SDMX Central (Test) will try to fix all semantic failures on conversion.

The dataset must conform to the structure as defined by the Data Structure Definition (DSD). The DSD defines the Dimensions and Attributes in a dataset. If the dataset is not structurally compliant then it indicates that these are either invalid Dimensions, or invalid Attributes i.e. the Dimension or Attribute Id in the dataset is not defined in the DSD.

The reported values in a dataset must conform to the allowable values are defined by the Data Structure Definition (DSD). Each component (dimension/attribute) defined by a DSD can additionally define an enumeration of allowed values (Codelist), or a component may provide restrictions on the content such as content type (String/Integer/Boolean), content length, etc. If any of the reported values do not conform to the values allowed by the DSD then the representation will be invalid.

The allowable content of a dataset may be further constrained by one or more Constraints. A Constraint can define allowed and/or restricted values for each data structure component. Constraints can also define allowed and/or restricted series/partial series.

Constraints can be applied to a Data Structure, Dataflow, Data Provider, or Provision Agreement. The rules of a constraint will cascade, so if a Constraint which restricts valid frequencies to Monthly or Annual is applied to a Dataflow, and another restricting valid countries to France is applied to a Data Provider, then the resultant restriction for the data provider loading data for the dataflow is "Monthly or Annual data for France".

The dataset may contain duplicate observations or series. The validation process will attempt to merge duplicates to create a consolidated dataset. If the merge process finds duplicates which are in conflict with each other, for example two observations for the same time period, reporting different observation values, then this will be reported as a duplicate error.

If Validation Rules exist for the structure defined by the data, then these will be evaluated against the data for correctness.

A Data Structure Definition (DSD) can define attributes which hold additional information about the dataset, series, or observation. An attribute may be optional or mandatory. The absence of mandatory attributes does not necessarily mean the dataset is invalid, as the dataset may be reporting additional information about an existing dataset, series, or observation. A dataset may also be used with a delete action, in which case only subsets of the dataset are required.

There are two types of constraint definitions, ones which define the allowable series keys (Series Constraint) and ones which define the allowable code ids for each Dimension and Attribute (Code Constraint).

---

Series Constraint

Selecting a Series Constraint will create a restriction based on all the series in the validated dataset.

Code Constraint

Selecting a Code Constraint will create a restriction based on all the unique code ids in the validated dataset.

The constraint id needs to be specified, so that the constraint can be uniquely identified.

---

A recommended naming convention is to use the text "CONS_" followed by either "KEY" or "CODE" depending on the constraint type, followed by the id of the structure that is being constrained. For example if the id of the constrained structure is "DF_GDP", the constraint id could be either "CONS_KEY_DF_GDP" or "CONS_CODE_DF_GDP"

If there is already a content constraint in the Fusion Metadata Registry with the same id, then the constraint action can be set to either overwrite this existing constraint with the new generated constraint, or to generate a new constraint and merge the allowable values with the existing constraint. A merged constraint will be the UNION of the constraint rules from both constraints, with no duplications of rules. The merged constraint will then be saved as a replacement of the existing constraint.

Note that 'Overwrite Existing' will not overwrite an existing constraint against a constraint of a different type.

The constraint structure is determined from the structure types that the dataset has been linked to from the dataset details page. The constraint will either be attached to a Provision Agreement, Dataflow, or Data Structure Definition, in that order. So if the dataset has been attached to a Dataflow but not a Provision Agreement, then the constraint will be attached to the Dataflow. If the dataset details are updated to attach the dataset to the Provision Agreement, then the constraint generation will update to attach the constraint to the Provision Agreement.

Constraint generated successfully. Details are:

Agency Id
Id
Version

To aid with data authoring an Excel template can be created. This will create an template in Excel format for the current Provision Agreement. Data can be entered into the template and then supplied back to the system to aid data creation.

Excel Reporting Templates enable data reporters to be provided with a custom built Excel template. The template provides a tabular form that the data provider can fill in with observation values. The form contains only the cells that the data provider is expected to report data for.

Use the value from the HTTP message header.

If no locale is specified either as a parameter in the URL or (and if the other toggle is set to Enabled) in the Accept-Language parameter of the header, then this value will be used.