Skip to main content

Assign Dataset Dimension

POST 
/datasets/:dataset_id/dimensions:assign

Save a row_id -> category assignment as a dataset-scoped dimension.

This is the user-authoring write path for dimensions: until now the only way a per-row dimension could be created was an analysis op run. The saved artifact is dataset-scoped, so it resolves as assigned:<name> on every later run of the same dataset — which is what makes categories comparable across runs.

Returns 422 if the name is malformed, the assignment is empty, it exceeds the row-count or category ceiling, or it references rows absent from the target version. Returns 409 if concurrent saves cannot be serialised.

Request​

Responses​

Successful Response