2025)¶

=====================================

🎉 Highlights¶

In this release, we're excited to add Windows support to Bodo. We also continue our focus on Iceberg and enhance our support for Iceberg Rest Catalogs.

✨ New Features¶

Bodo is now available on Windows on pip and Conda.
BodoSQL's TabularCatalog has become IcebergRestCatalog.
Test suites for Iceberg Rest Catalog using Polaris.

🐛 Bug Fixes¶

Fixed a hang when using spawn mode from an interactive Python session.
Fixed an incorrect output array type when rewriting free variables to arguments in UDFs

2025.3.1¶

🎉 Highlights¶

In this release, we introduce the map_parititions API, extend our Parquet I/O to support Multi-Index, and upgrade to the latest Arrow and Pyiceberg!

✨ New Features¶

Added BodoDataFrame.map_partitions API, which applies a function in parallel over partitions of the dataframe.
Extend our Parquet read and write to support Multi-Index for Pandas DataFrames.

⚙️ Dependency Upgrades¶

Upgraded to Pyiceberg 0.9.
Upgraded to Arrow 19.
Removed bodo-azure-sas-token-provider, Bodo uses Arrow for Azure IO.

2025.3.2¶

Minor bugfix release. - Fixed bug in to_parquet when running from cache. - Fixed outputs for jupyter notebooks with spawn mode when the kernel is run separately from the notebook server.