Hyperleaup 0.1.2

We’d like to announce the release of Hyperleaup 0.1.2, which introduces the following improvements:

• Adds an experimental LARGEFILE creation mode option to improve performance when creating a Hyper file for a large dataset. It does this by saving multiple files and using an array of Parquet file paths in the Tableau Hyper API COPY command.

Note: This creation mode only works with the Databricks File System (DBFS).

Credits: Dustin Vannoy (@dustinvannoy-db)

Hyperleaup 0.1.1

We’d like to announce the release of Hyperleaup 0.1.1, which introduces the following improvements:

• Adds a new timestamp_with_timezone option to the HyperFile and Creator classes. This option will default to False to keep backward compatibility, but setting this to True will avoid the common error:

HyperException: Type 'timestamp' is incompatible to type 'timestamptz' of column 'last_updated' in Parquet file.

• Adds a new allow_nulls parameter to HyperFile class which replaces null numbers in the input dataset to 0.

This release can be installed using pip:

pip install hyperleaup==0.1.1

Credits: Dustin Vannoy (@dustinvannoy-db)

Hyperleaup 0.1.0

We’d like to announce the release of Hyperleaup 0.1.0, which introduces the following improvements:

• Project restructuring in accordance with the official Python Packaging User Guide
• Project packaging using Hatchling
• Bump PySpark version from 3.1.3 to 3.2.2

You can now download and install Hyperleaup directly from PyPI: pip install hyperleaup

Hyperleaup 0.1.0-beta

Hyperleaup 0.1.0-beta
This (beta) release adds the following features and bug fixes:

• Adds support for Tableau server token authentication
• Bumps the version of Pyspark from 3.1.2 to 3.1.3
• Adds an Apache 2.0 license

Hyperleaup 0.0.6

Hyperleaup release 0.0.6
This initial release supports creating Tableau Hyper files from Apache Spark DataFrames.