Cumulative Distribution

Concepts: sqlCumulativeDist

Distribution Functions Two functions express a row's position as a decimal between 0 and 1. Understanding their subtle differences is key to choosing the right one. CUME_DIST PERCENT_RANK Distribution Patterns Distribution functions power some of the most valuable analytics patterns: outlier detection, sessionization, and percentile-based tiering. CUME_DIST Anomaly Detection This query identifies the top 1% of transactions by amount. These are the outliers that warrant manual review. The same pattern works for detecting unusually long response times, abnormally high event counts, or any metric where extreme values signal problems. Sessionization Pattern Sessionization assigns a session identifier to groups of events that are close together in time. The standard approach detects gaps greate