Using the API
CSV and bulk downloads
JSON is the default, but you don't have to use it. Ask for CSV to open results in a spreadsheet, or download a whole dataset in a single request.
Choosing a format
Endpoints that offer more than one format take a format parameter:
| Format | Best for |
|---|---|
json | Apps and scripts. The default. |
csv | Spreadsheets, and tools like pandas or R. |
ndjson | Streaming, with one JSON record per line. |
curl "https://api.ukdataatlas.com/v1/data/population?period=2022&geography=ward&boundaryYear=2023&format=csv"The API picks the format from this parameter only, not the Accept header.
What CSV looks like
- Each row carries its own provenance, so a saved file still says where it came from.
- Rows include the measure's
unit, so values can't be misread. lowerBoundandupperBoundcolumns hold any published confidence interval, and are empty otherwise.- In lookup tables, a list is joined with
|in a single cell.
Whole datasets
Paging through a big dataset takes many requests. Instead, list the downloads on offer and fetch the one you want as a single file:
curl "https://api.ukdataatlas.com/v1/exports"Each download has a contentHash. Check it after downloading to be sure you got exactly the file that was published. Keep a copy in sync shows how.
/exports/{export-id}Download a datasetLookup tables
Lookup tables are the reference data you join against: area codes and names for each boundary release, crosswalks as one row per mapping, and named-location membership. Download them as CSV or NDJSON:
curl "https://api.ukdataatlas.com/v1/lookups/areas-combinedAuthority-2025-12-en-bgc?format=csv"Map tiles
Boundaries for maps come as vector tiles, or as one PMTiles archive holding every tile for a release. See Draw a map.