Databricks SQL adds MATCH_RECOGNIZE operator for sequence pattern detection
bloganalyticsdatabrickspreviewengineer
feature
Databricks SQL now includes the `MATCH_RECOGNIZE` operator, available in Public Preview, which introduces regex-like pattern matching for rows. This new operator simplifies the detection of sequences and patterns within event data, a task often challenging with standard SQL. It allows users to express complex temporal and sequential logic directly in a single SQL clause, making it easier to identify trends in cybersecurity logs, financial data, e-commerce funnels, and IoT sensor streams. The functionality is available across Databricks compute environments, including Lakehouse Real-Time.
Features (1) ›
- Added the item to their cart (ADD_TO_CART)
Read the original announcement →
https://www.databricks.com/blog/regex-rows-simplifying-pattern-detection-sql-matchrecognize
Related releases
- Databricks Lakeflow Jobs to Support JAR Tasks on Serverless Compute for Compliance Profiles Databricks Release Notes ·
- Databricks Apps Enabled by Default for Compliance Workspaces Databricks Release Notes ·
- Genie Code adds conversation controls for branching and editing chats Databricks Release Notes ·
- Databricks SQL Adds counter_diff Window Function for Time-Series Analysis Databricks Release Notes ·
- Databricks Rolls Out Automatic Change Data Feed for Delta Lake and Apache Iceberg Databricks Release Notes ·
- Databricks Go SDK v0.180.0 Adds GroupName to Pipelines.RunAs Configuration Databricks Go SDK Releases ·