Search results

Add explicit jax kernel tests

testsMore or better tests are neededMore or better tests are needed

Status: Open.

#3466 In scikit-hep/awkward;

· ikrommyd opened on Apr 14, 2025

ak.any yields True for empty rows in float arrays with the jax backend

bugThe problem described is something that must be fixedThe problem described is something that must be fixed

Status: Open.

#3465 In scikit-hep/awkward;

· ikrommyd opened on Apr 14, 2025

ak.argmin/argmax give incorrect results in the jax backend

bugThe problem described is something that must be fixedThe problem described is something that must be fixed

Status: Open.

#3463 In scikit-hep/awkward;

· ikrommyd opened on Apr 13, 2025

All jax backend reducers give wrong length output if the last element of the input jagge arrays is empty

bugThe problem described is something that must be fixedThe problem described is something that must be fixed

Status: Open.

#3462 In scikit-hep/awkward;

· ikrommyd opened on Apr 13, 2025

Strange segmentation fault of awkward_IndexedArray64_numnull with jax and cuda installation

bug (unverified)The problem described would be a bug, but needs to be triagedThe problem described would be a bug, but needs to be triaged

Status: Open.

#3458 In scikit-hep/awkward;

· ikrommyd opened on Apr 12, 2025

ak.min/max/argmin/argmax do not work on axis=1 for the jax backend if the last element of an input ListOffsetArray is empty

bugThe problem described is something that must be fixedThe problem described is something that must be fixed

Status: Open.

#3456 In scikit-hep/awkward;

· ikrommyd opened on Apr 11, 2025

Implement consistent __setitem__ between numpy/cupy.ndarrays and VirtualArrays

Status: Open.

#3444 In scikit-hep/awkward;

· ikrommyd opened on Apr 6, 2025

Add VirtualArray in the input/output signature of the nplikes' functions when required

Status: Open.

#3429 In scikit-hep/awkward;

· ianna opened on Mar 19, 2025

Investigate the usage of cached_property for materialize and is_materialized in VirtualArray

Status: Open.

#3428 In scikit-hep/awkward;

· ikrommyd opened on Mar 19, 2025

Should the generator of a VirtualArray always yield unique copies of an array?

Status: Open.

#3422 In scikit-hep/awkward;

· pfackeldey opened on Mar 18, 2025

Should VirtualArrays allow JAX as nplike?

Status: Open.

#3421 In scikit-hep/awkward;

· pfackeldey opened on Mar 18, 2025

update _backends so that we don't have to hard-code the nplike-to-backend relationship

Status: Open.

#3420 In scikit-hep/awkward;

· ianna opened on Mar 18, 2025