For data analytics, 19c improves the APPROX_COUNT function set. The PDF documentation details how to use these functions to provide instant approximate answers for distinct counts
| Feature Area | Description | |--------------|-------------| | | Native SQL/JSON functions ( JSON_OBJECT , JSON_ARRAY , JSON_TABLE , JSON_EXISTS ). | | SQL Polymorphism | SQL_MACRO (scalar and table macros) for reusable SQL logic. | | Private Temporary Tables | Session-specific temporary tables that auto-cleanup. | | Real-Time Statistics | Automatic online statistics refresh for bulk DML operations. | | Automatic Indexing | AI-managed indexes (recommendation and creation). | | Hybrid Partitioned Tables | Combine internal partitions with external data sources. | | RESTful Services | Oracle REST Data Services (ORDS) integrated with SQL. | | Approximate Queries | APPROX_COUNT_DISTINCT , APPROX_PERCENTILE for faster big data analytics. | | Multi-tenant Enhancements | CONTAINERS() clause to query across all PDBs. | | In-Memory SQL | In-Memory Column Store (IMCS) with optimized vector joins. | oracle sql 19c pdf
-- JSON generation from relational table SELECT JSON_OBJECT('dept_id' IS d.department_id, 'dept_name' IS d.department_name) AS dept_json FROM departments d; For data analytics, 19c improves the APPROX_COUNT function