Validation Rule Reference Information

Overview

Reference information about validation rules (DQ Rules).

Rule Types

Category	Rule Type
Foundational DQ	DATA FORMAT
	BLANKS
	VALUERANGE
	UNIQUENESS
	UNIQUE VALUES
	DATA ANALYSIS EXCEPTIONS
	CUSTOM QUERY
Pipeline Checks	COUNT
	FRESHNESS
	METRIC
Cross-System DQ	ORPHAN RECORDS
	ORPHAN KEYS
	INTEGRITY CHECK
MDM	ADDRESS VALIDATION
	DOC-MATCHING
Data Diffs	COMPARE COUNT
	COMPARE METRIC
	COMPARE TABLE
	COMPARE CUSTOM QUERY

Run Status

A new rule will have an implicit Run Status of Valid. After the rule is executed, there are a variety of Run Results that the rule may take. "Run Status" and "Run Result" are used interchangeably in DvSum DQ.

Status	Icon	Definition
Passed		Rule is valid, and no exceptions were found in the data in the most recent run.
Failed		Rule is valid, and exceptions were found in the most recent run. This applies only to rules which do not identify specific invalid records. For example, a COUNT rule result falls outside of the allowed range.
Exception		Rule is valid, and at least one exception was found in the data in the most recent run. This applies to rules which identify invalid records. For example, a failed BLANKS or UNIQUENESS check will result in this status.
Invalid		Rule is not valid. This does not indicate a problem with the data. Rather, it indicates a problem with the rule definition which must be solved before the rule can be executed.
Matched		For example, a ADDRESS VALIDATION will indicate Matched after successful execution.
Modified		After data steward performs cleansing, but data is not yet committed back to source.
Committed
Valid

Run Status & Readiness Score

Along with a status, a Readiness Score is also generated. Readiness score provides a second degree of information on the quality of data. If the data rule is failing, how bad is it. Readiness score creates a common unit for measuring the quality of data and allows result of audits calculating the overall data quality score across various data elements and types of audits.

Sample Audit Result	Audit Result Icon		Readiness Score
Count of finished good items in item master is 6000 which is within the tolerance of 5500 and 7000			100%
Count of finished good items in item master is 5000 which is not within the tolerance of 5500 and 7000			91% Explanation: 5000 is 9% short of 5500. Calculation: 1 - (500/5500)
Forecast Name is 100% unique in SALES_FORECAST extract			100%
Run over run variance of qtyplanned in PURCHASE_PLAN is 5% for Supplier X which is more than the limit of 3%			33% Calculation: 1 - (.02/.03)
There are 2 routing records where yield is not between 0 and 1. (total records are 100)			98% Calculation: 1 - (2/100)

 Following insights are available with your audit results

Icon	Insight Type	Definition
	History	History for all audits
	Exception List	List of Exceptions for master data audits
	Drill-down Analytical	Drill-down with Variance for aggregation audits

History Trend Insights

With history, you can get insights into the changes to the audit result from last run and also the trend of that audit over time. History insights can be useful to identify what audits to focus on.

History Trend	Definition	Focus Required
	Flat. No change	Status Quo
or	Positive Up – Test was a fail before and now pass. Test has been failing because results too low but now trending up. Positive Down – Test has been failing because results were too high and now trending down.	Things are improving even if the tests are failing. May not need to focus.
or	Negative Down – Test was passing before and is now failing. Or it has been passing but values are trending down and will cross the lower tolerance soon. Or it has been failing and things are getting worse Negative Up – Test was passing before and is not failing. Or it has been passing but values are trending up to cross the upper tolerance soon	Things are getting worse for failed tests – requires top priority focus. Things are trending in the wrong direction for passing tests – heads-up for future issues