v3.16.1

What's Changed

🦠 Bugfixes

StreamingDataFrame: retain a custom stream_id across operations by @daniil-quix in #925

⚠️ Note: This fix may affect the existing stateful operations on the DataFrames with single-partition topics called after .group_by() due to optimizations introduced in v3.14.0.

The stream_id is used as part of the State stores' names, and it wasn't propagated correctly, leading to incorrect store names in some cases.
The fix in #925 corrects that, but the state stores created after .filter() or .apply() operations on the grouped DataFrame won't be accessible anymore because of the corrected stream ids.
See #925 for more examples of affected code.

Other fixes

• Fix conda requirements by @daniil-quix in #920
• sinks/InfluxDB3: adjust check to work with v2 and v3 by @tim-quix in #922

Full Changelog: v3.16.0...v3.16.1