0.9.1

Fixes relative timestamp behavior for Pandas >=3.

A change in Pandas caused relative timestamps like -2h to be parsed as unexpected absolute timestamps causing queries to fail against the backend.

Thank you, Greg Anderson for bringing this to our attention.

What's Changed

• Fix resolve_time to be Pandas 3 compatible by @seanshahkarami in #14

Full Changelog: 0.9.0...0.9.1

Release 0.9.0

Adds support for new experimental aggregation window in Sage Data API via new "experimental_window" param in query. This param accepts duration values such as "1h", "8h", and "1d".

Release 0.8.0 Post 1

Added documentation for installing in a Python virtual environment.

Release 0.8.0

Adds support for new experimental aggregation functions in Sage Data API via new "experimental_func" param in query. This param accepts values "count", "sum", "mean", "min", "max.

Release 0.7.1

Falling back to standard Python json module for now to allow both very large / small numbers and mixed value types.

Release 0.7.0

Please do not use this version! The switch to pyarrow has broken support for mixed type values! We will address this ASAP!

• Use pyarrow engine internally to support very large and small numbers when loading data.
• Bumped minimum pandas version to 2.0.0.

Release 0.6.0

Added support for gzip compression when querying data.

Release 0.5.0 Post 1

Updated query parameter reference in docs.

Release 0.5.0

Added head parameter to query to allow querying the earliest head records in each group found.

Release 0.4.2 Post 1

Added note for contributors.