[renovate]

This PR contains the following updates:

Package	Type	Update	Change
sparse	feature-backends	patch	>=0.19.2 → >=0.19.3

---

Release Notes

pydata/sparse (sparse)

v0.19.3: Version 0.19.3

Compare Source

What's Changed

• [pre-commit.ci] pre-commit autoupdate by @​pre-commit-ci[bot] in #​952
• fix: support negative offsets, rectangular shapes, and negative axes in sparse.diagonal by @​rautaditya2606 in #​955
• [pre-commit.ci] pre-commit autoupdate by @​pre-commit-ci[bot] in #​953
• [pre-commit.ci] pre-commit autoupdate by @​pre-commit-ci[bot] in #​956
• fix: avoid dense column selectors in GCXS basic slicing by @​Doribelove in #​957
• fix(io): support CSR, CSC, and DOK formats in save_npz and load_npz by @​rautaditya2606 in #​960
• DOC: name the array parameter x in expand_dims and flip by @​hxperl in #​958
• [pre-commit.ci] pre-commit autoupdate by @​pre-commit-ci[bot] in #​961
• fix: preserve zero-dimensional arrays in sparse conversions by @​Doribelove in #​963
• fix: GCXS matmul with zero output columns by @​raashish1601 in #​972
• fix: return the identity for all/any over an empty axis by @​raashish1601 in #​971
• fix: honour dtype in asarray for sparse inputs by @​raashish1601 in #​975
• fix: sort unique_counts output when the fill value is inserted by @​raashish1601 in #​974
• fix: keep the product dtype in 1-D dot and matmul by @​raashish1601 in #​973
• docs: explain GCXS on the front page by @​Doribelove in #​968
• docs: clarify COO idx_dtype capacity for from_numpy by @​Doribelove in #​967
• feat: support Fortran order in COO.flatten by @​Doribelove in #​964
• fix: normalize negative axis in argmax and argmin by @​Rodrigo-Palma in #​966
• fix: accept GCXS input in triu, tril and diagonal by @​raashish1601 in #​978
• fix: sort NaNs to the end in sparse.sort by @​raashish1601 in #​979
• fix: allow any and all on arrays with a non-boolean fill value by @​raashish1601 in #​980
• fix: return the first NaN in argmax and argmin by @​raashish1601 in #​976
• fix: support ufuncs with several outputs such as divmod, modf and frexp by @​raashish1601 in #​982
• fix: broadcast array bounds in clip by @​raashish1601 in #​981
• fix: skip complex NaNs in nansum and nanprod by @​CAOShurong in #​988
• docs: explain fill-value propagation in where by @​Doribelove in #​987
• fix: index GCXS arrays pointwise with several index arrays by @​raashish1601 in #​985
• fix: move the advanced-index axis to the front like NumPy when advanced indices are separated by @​raashish1601 in #​984
• fix: put new axes in the right place when indexing GCXS after an integer index by @​raashish1601 in #​983
• Add validation and clean up compressed_axes=None ambiguity in GCXS. by @​hameerabbasi in #​1015

New Contributors

• @​rautaditya2606 made their first contribution in #​955
• @​Doribelove made their first contribution in #​957
• @​hxperl made their first contribution in #​958
• @​raashish1601 made their first contribution in #​972
• @​Rodrigo-Palma made their first contribution in #​966
• @​CAOShurong made their first contribution in #​988

Full Changelog: pydata/sparse@0.19.2...0.19.3

---

Configuration

📅 Schedule: (UTC)

• Branch creation
  • At any time (no schedule defined)
• Automerge
  • At any time (no schedule defined)

🚦 Automerge: Enabled.

♻ Rebasing: Whenever PR is behind base branch, or you tick the rebase/retry checkbox.

🔕 Ignore: Close this PR and you won't be reminded about this update again.

---

• If you want to rebase/retry this PR, check this box

---

This PR was generated by Mend Renovate. View the repository job log.