Skip to main content

Retire entities

We only retire entities which have been added in error.

Entities which are no longer in use, for example where a nature reserve is de-designated, a world heritage site is de-listed, a site has been developed and removed from the brownfield-land register, or a listed building has been de-listed should continue to be in the dataset marked with an end-date added.
The end-date should be ideally added to the data source the data provider, but may be provided in a secondary source provided by us.

Where two or more entities exist for the same entity, they should be merged rather than deleting the duplicate entity.

An entity which has been added in error can be removed by adding it to the old-entity config.

  1. Identify the entity number
    First, you need to identify the entity number of the entity you wish to remove. This can easily be done by filtering by the reference in lookup on datasette. The link is filtering for the reference A4Da1d as an example, just replace it with the appropriate reference. Copy the entity number under entity.

  2. Populate old-entity.csv

    Next, add a new row to the old-entity.csv file located in the corresponding pipeline folder E.g., if the dataset is article-4-direction, populate the old-entity.csv file within that pipeline folder. Add the corresponding entity number, the related HTTP response error code (this should be 410 if the entity is being retired). You should also add a comment with the reason for its deletion in the notes column.

For example, the populated new line could look like this:

7010002871,410,,new endpoint and reference field for GLO,,,

Once done, push the changes.