What is a Conditional metric?
Conditional metrics define row-level quality rules that every record in an asset must satisfy. They answer the question: Do the values in this dataset meet the business conditions we require? Type: User-defined — Conditional metrics are created manually. Navigate to Metric → Add and select Subcategory: Conditional. You can optionally scope the rule to a subset of rows using Select criteria. When no criteria is set, the rule applies to all records in the asset. Prizm scores the metric as(Valid records / Total records) × 100. Records that fail the condition rules count as invalid and contribute to alert generation and the asset’s overall quality score.
Metrics
Setup
Step-by-step guide to creating and configuring a conditional metric.
Use Cases
Real-world examples — range enforcement, completeness, cross-column consistency.