Changelog

Changelog#

v0.1.3 (2025-03-13)#

Bug Fixes#

  • Fix link to cuDF documentation in the README [#7]

  • Update typehint of Device.parse method to accept torch.device objects [#8]

  • Add missing py.typed file to the package [#9]

v0.1.2 (2025-03-10)#

Refactoring#

  • Use relative imports in the source code [#6]

Bug Fixes#

  • Add data type (dtype) assertion checks in the testing functions in abracudabra.testing module [#5]

  • Fix the data type conversion issues between torch and CuPy/cuDF objects [#5]

v0.1.1 (2025-03-09)#

Documentation#

  • Improve the doctrings for all public functions

Sphinx Documentation#

  • Move typehints to the description

  • Hide private functions

  • Add missing version switcher

Refactoring#

  • Rename the annotations module to _annotations (private module)

Bug Fixes#

  • Fix the ValueError in guess_device function, that happens when no device is found because the provided sequences are not NumPy/CuPy/PyTorch/Pandas/cuDF valid objects.

v0.1.0 (2025-03-09)#

This is the first release of the project. It includes the following features:

  • Data conversion: to_array, to_tensor, to_dataframe, to_series

  • Device management: to_device, get_device

  • Library management: get_np_or_cp, get_pd_or_cudf

Contents