Release 0.182#
General#
- Fix correctness issue that causes - corr()to return positive numbers for inverse correlations.
- Fix the EXPLAIN query plan for tables that are partitioned on - TIMESTAMPor- DATEcolumns.
- Fix query failure when using certain window functions that take arrays or maps as arguments (e.g., - approx_percentile()).
- Implement subtraction for all - TIMEand- TIMESTAMPtypes.
- Improve planning performance for queries that join multiple tables with a large number columns. 
- Improve the performance of joins with only non-equality conditions by using a nested loops join instead of a hash join. 
- Improve the performance of casting from - JSONto- ARRAYor- MAPtypes.
- Add a new IPADDRESS type to represent IP addresses. 
- Add - to_milliseconds()function to convert intervals (day to second) to milliseconds.
- Add support for column aliases in - CREATE TABLE ASstatements.
- Add a config option to reject queries during cluster initialization. Queries are rejected if the active worker count is less than the - query-manager.initialization-required-workersproperty while the coordinator has been running for less than- query-manager.initialization-timeout.
- Add TPCDS connector. This connector provides a set of schemas to support the TPC Benchmark⢠DS (TPC-DS). 
CLI#
- Fix an issue that would sometimes prevent queries from being cancelled when exiting from the pager. 
Hive#
- Fix reading decimal values in the optimized Parquet reader when they are backed by the - int32or- int64types.
- Add a new experimental ORC writer implementation optimized for Presto. We have some upcoming improvements, so we recommend waiting a few releases before using this in production. The new writer can be enabled with the - hive.orc.optimized-writer.enabledconfiguration property or with the- orc_optimized_writer_enabledsession property.