v3.0.0

3.0.0 (2022-03-29)

⚠ BREAKING CHANGES

• BigQuery Storage and pyarrow are required dependencies (#776)
• use nullable Int64 and boolean dtypes in to_dataframe (#786)
• destination tables are no-longer removed by create_job (#891)
• In to_dataframe, use dbdate and dbtime dtypes from db-dtypes package for BigQuery DATE and TIME columns (#972)
• automatically convert out-of-bounds dates in to_dataframe, remove date_as_object argument (#972)
• mark the package as type-checked (#1058)
• default to DATETIME type when loading timezone-naive datetimes from Pandas (#1061)
• remove out-of-date BigQuery ML protocol buffers (#1178)

Features

• add api_method parameter to Client.query to select INSERT or QUERY API (#967) (76d88fb)
• default to DATETIME type when loading timezone-naive datetimes from Pandas (#1061) (76d88fb)
• destination tables are no-longer removed by create_job (#891) (76d88fb)
• In to_dataframe, use dbdate and dbtime dtypes from db-dtypes package for BigQuery DATE and TIME columns (#972) (76d88fb)
• mark the package as type-checked (#1058) (76d88fb)
• use StandardSqlField class for Model.feature_columns and Model.label_columns (#1117) (76d88fb)

Bug Fixes

• automatically convert out-of-bounds dates in to_dataframe, remove date_as_object argument (#972) (76d88fb)
• improve type annotations for mypy validation (#1081) (76d88fb)
• remove out-of-date BigQuery ML protocol buffers (#1178) (76d88fb)
• use nullable Int64 and boolean dtypes in to_dataframe (#786) (76d88fb)

Documentation

• Add migration guide from version 2.x to 3.x (#1027) (76d88fb)

Dependencies

• BigQuery Storage and pyarrow are required dependencies (#776) (76d88fb)