View a markdown version of this page

Run rulesets and view results - Amazon SageMaker Unified Studio
Run a rulesetView run resultsView evaluation run detailsReview the dashboard

Run rulesets and view results

After you create a ruleset, you can run it on demand or on a schedule and view the results on the Data quality tab.

Run a ruleset

  1. On the Rulesets sub-tab, find the ruleset you want to evaluate.

  2. Choose Run. You have two options:

    • Run – Evaluate this ruleset against the entire table.

    • Run on a subset – Filter the table with a preprocessing SQL query, then evaluate only the matching rows.

    Run dropdown showing Run and Run on a subset options

  3. If you choose Run on a subset, enter a SELECT statement in the Preprocessing query editor. Only SELECT statements are supported. The query filters the table data for this run only and does not change the ruleset.

    Run on a subset dialog with preprocessing query editor

  4. Choose Run to start the evaluation.

  5. When the run completes, the ruleset card updates with the results showing which rules passed and which failed. You can view the SQL query used for the run under Evaluation scope in the results and in run details.

For more information about preprocessing queries, see Using preprocessing queries in the AWS Glue User Guide.

View run results

After a run completes, the ruleset card shows the status (Passed or Failed), score (percentage of rules that passed), rules passed (for example, "1 of 2"), and Started On and Completed On timestamps.

Expand the ruleset card to see a detailed table with each evaluated rule.

Ruleset with expanded run results

The results table shows:

Column Description
Evaluated rule The DQDL rule expression that was evaluated.
Evaluated metrics The measured value (for example, row count of 10,002).
Evaluation message Details on why a rule failed.

You can search and filter the results using the search box and status filter above the table.

View evaluation run details

Select a run from the Rulesets sub-tab or Run history sub-tab to open the Evaluation Run Details page.

Evaluation Run Details page

The details page shows:

Field Description
Ruleset The ruleset that was evaluated.
Run ID Unique identifier for this run.
Execution Status Whether the run completed successfully.
Started On Timestamp when the evaluation began.
Completed On Timestamp when the evaluation finished.
Log group CloudWatch Logs log group for detailed execution logs.
Quality result Overall result (Passed or Failed).
Quality score Percentage of rules that passed.

Below the summary, three tabs provide additional detail:

  • Output Logs: Standard output from the evaluation run.

  • Error Logs: Any errors that occurred during the run.

  • Rule Results: A table showing each evaluated rule, its measured metrics, and evaluation messages.

Review the dashboard

After you have run at least one ruleset, the Data quality tab shows a dashboard with summary metrics at the top, followed by three sub-tabs: Rulesets, Rule results, and Run history.

Data quality tab with dashboard

The dashboard tiles show:

Tile Description
Quality score The average of the latest quality scores per ruleset.
Rulesets Number of rulesets and total active rules across all sets on this table.
Execution success rate Percentage of runs that completed without errors.
Last run Timestamp, ruleset name, and status of the most recent evaluation.
Score trend Quality score over time, showing how data quality changes between runs.

Use the filters above the sub-tabs to focus results:

  • Filter by ruleset: View results for a specific ruleset when you have multiple.

  • Status: Filter by All, Passed, or Failed.

  • Time range: Filter by relative time window or absolute date range.