dr_quality()

An asset selects its latest attempt, while release selects the evidence for that exact published stand. Cached runs resolve to the original release's checks. These choices prevent an older successful run hiding a recent failure.

Usage

dr_quality(
  x,
  run_id = NULL,
  asset = NULL,
  release = NULL,
  ...,
  action = "block",
  threshold = 0,
  engine = "native",
  volatile = FALSE
)

Arguments

x: A one-sided formula to construct a quality rule, or a lake, run result, measurement, measurement set, dbt result or quality tibble. Measurement sets resolve their exact input releases using retained references. Unavailable references return explicit not_checked evidence.
run_id: Run ID when x is a lake.
asset: Asset ID. Required with release; otherwise selects the latest attempt. Supply at most one of run_id and asset.
release: Exact release ID to inspect, together with asset.
...: For a rule, optional name, description and dimension.
action, threshold, engine, volatile: See dr_quality_rule(). New definitions use only action and threshold as publication policy.

Value

A quality tibble. failure_rate is derived from available counts. Native pointblank thresholds are recorded in the JSON details column.

Examples

contract <- dr_contract("orders", version = "1",
  columns = c(id = "integer"), key = "id")
dr_quality(dr_validate(data.frame(id = c(1L, 1L)), contract))

dr_quality()

Usage

Arguments

Value

Examples

See also

Search documentation