Hello, Pandas has Iceberg support merged (using PyIceberg in the backend), which will be available in the upcoming Pandas 3 release. Any feedback from the Iceberg community can be very helpful at this initial stage. But please keep in mind that the goal is to democratize Iceberg for Python developers and not provide the most complex features right away. Here are the PRs:
https://github.com/pandas-dev/pandas/pull/61383 https://github.com/pandas-dev/pandas/pull/61507 One challenge is that specifying partitioning for Iceberg write in DataFrame.to_iceberg requires PyIceberg objects in the API signature, but making PyIceberg a hard dependency is not acceptable for Pandas. Any thoughts on other ways to specify partitioning in Python APIs? BTW, Bodo has released an implementation of these Pandas APIs if you want to play with them now quickly (since Pandas is not released yet): https://docs.bodo.ai/latest/quick_start/quickstart_local_iceberg/ Best, Ehsan
