• Files Changed
• Browse Source

- update to 2.7.1:
  * fix: test from_raggedtensor on CUDA
  * fix: use None when ndims can't be inferred from a layout-like
    obj
  * fix: ak.argcombinations should allow negative axis
  * fix: restrict `named_axis` inferring to
    `ak.Arrays/ak.Records/ak.HighLevelContexts` by default
- update to 2.7.0:
  * chore: remove Python 3.8, ensure Python 3.13 is included
- update to 2.6.10:
  * feat: named axis for `ak.Array`
  * feat: to/from TensorFlow Tensor
  * feat: updated LayoutBuilder.h (added functors) for C++11
    compatibility
  * fix: make sure 'at' is a cupy zero dim array
  * fix: add cuda backend support for `to_raggedtensor` and
    `from_raggedtensor` functions
  * fix: test `from_raggedtensor` on GPU
  * fix: correct handling of keepdims and mask_identity for
    weighted mean
- update to 2.6.9:
  * feat: Add to_cudf
  * feat: to/from PyTorch Tensor
  * perf: avoid inflating UnmaskedArrays in broadcasting when you
    can
  * fix: TypeError fix for `can_cast`
  * fix: ListArray slicing on GPU
  * fix: ak.typetracer.length_one_if_typetracer with option and
    union types

• Files Changed
• Browse Source

baserev update by copy to link target

• Files Changed
• Browse Source

- Revert numpy restriction to > 2
  * Upstream declares 1.18.0.
- Test both in multibuild:
  * numba: requires numpy < 2.1 (--> installs numpy1)
  * no numba --> installs numpy >= 2.1
- enable uproot tests: no cycle in multibuild

• Files Changed
• Browse Source

Update to version 2.6.8

• Files Changed
• Browse Source

baserev update by copy to link target

• Files Changed
• Browse Source

- update to 2.6.5:
  * feat: 2772 pyarrow doesnt permit selective reading with
    extensionarray
  * fix: skip failing test when jax is not installed
  * fix: skip tests that require pyarrow when it is not installed
  * fix: generate array view
  * fix: Header-only fix is_valid for empty Indexed or
    IndexedOption
  * fix: wait for Jitify performing a one-time only warm-up
  * fix: prevent exponential memory growth in UnionArray

• Files Changed
• Browse Source

baserev update by copy to link target

• Files Changed
• Browse Source

- update to 2.6.4:
  * feat: `TypeTracerReport.data_touched_in` should ignore
    unrecognized `node_ids`
  * feat: better error message for missing dispatch
  * feat: add boolean cases for CPU kernels
  * feat: add ak.flatten and ak.concatenate kernels
  * fix: restrict ak.softmax to axis=-1, the only axis for which
    the calculation is correct right now
  * fix: Numpy 2.0 out of bounds error for BitMaskedArray Py
    kernels
  * fix: _is_getitem_at_placeholder logic was backward for
    UnionArray
  * fix: refuse dak.Array in hidden-but-public-functions
  * fix: PyPy segfault may be related to using out-of-scope array
  * fix: distributed rdataframe is not yet support
  * fix: support conversion to arrow and back with non-option
    Unknown type
  * fix: intermittent segfault in PyPy
  * fix: is_valid checks in header-only library
  * fix: UnmaskedArray was missing its _nextcarry_outindex

• Files Changed
• Browse Source

baserev update by copy to link target

• Files Changed
• Browse Source

> Update to version 2.6.3.
> Relax awkward-cpp requirement to allow any awkward-cpp version above the minimum specified for building awkward.
> Disable a test that depends on jax/jaxlib.

• Files Changed
• Browse Source

baserev update by copy to link target

• Files Changed
• Browse Source

- Update to 2.6.2:
  * New features
    + feat: add CUDA kernels that calculate length/sum
    + feat: add variable length loop kernels
    + feat: make ak.mean and ak.count differentiable
  * Bug-fixes and performance
    + fix: remove redundant(?) Jax.is_tracer_type check in _layout
    + fix: throw more meaningful error message for non-NumPy buffers
    + fix: proper typetracer array slicing in BitMaskedArray
    + fix: ak.flatten on BitMaskedArray
    + fix: performance of ufunc resolution for non-nominal signatures

• Files Changed
• Browse Source

baserev update by copy to link target

• Files Changed
• Browse Source

- Update to version 2.6.1:
  The purpose of this release is to provide a lower-bound on the
  fsspec dependency version (#3006).
- 2.6.0:
  From this point onward, awkward strictly depends on fsspec.
  * feat: turn on CUDA unit tests for working kernels and add some CUDA kernels
    (gh#scikit-hep/awkward#2930)
  * feat: add string builder
    (gh#scikit-hep/awkward#2899)
  * feat: add to_parquet_row_groups
    (gh#scikit-hep/awkward#2979)
  * feat!: array is no longer allowed on NumpyArray and EmptyArray
    (gh#scikit-hep/awkward#2997)
  * fix: include header-only files in sdist
    (gh#scikit-hep/awkward#2909)
  * fix: touching of unions
    (gh#scikit-hep/awkward#2906)
  * fix: lazily project union contents
    (gh#scikit-hep/awkward#2913)
  * fix: support typetracer in to_packed for IndexedOptionArray
    (gh#scikit-hep/awkward#2912)
  * fix: cartesian nested record
    (gh#scikit-hep/awkward#2929)
  * fix: cast bool_ to int8 for positional reducers
    (gh#scikit-hep/awkward#2935)
  * fix: ak.std and ak.var for axis != -1.
    (gh#scikit-hep/awkward#2918)

• Files Changed
• Browse Source

baserev update by copy to link target

• Files Changed
• Browse Source

Note: Depends on sr#1138806.

* Update to version 2.5.2.
* Bump required version of awkward-cpp to 28.

• Files Changed
• Browse Source

baserev update by copy to link target

• Files Changed
• Browse Source

- Update to version 2.5.1:
- Bump required version of python-awkward-cpp to 27.

• Files Changed
• Browse Source

baserev update by copy to link target

• Files Changed
• Browse Source