Feature sparse linalg solvers - #2841

Open

abagusetty wants to merge 162 commits into

IntelPython:masterfrom

abagusetty:feature-sparse-linalg-solvers

Commits on Apr 2, 2026

Commits on Apr 3, 2026

Fix deprecated tol kwarg in SciPy host fallback
abagusetty
committed
2384185
sparse/linalg: complete LinearOperator algebra, CG/GMRES/MINRES with oneMKL hooks
abagusetty
committed
4720291

Commits on Apr 6, 2026

Add tests for scipy.sparse.linalg: LinearOperator, cg, gmres, minres
abagusetty
committed
58cc44b
Fix implicit numpy conversion; use .asnumpy() for dpnp arrays
abagusetty
committed
3a50062
tests: add comprehensive sparse linalg tests for LinearOperator, cg, gmres, minres
abagusetty
committed
62cf7a4
tests: rewrite sparse linalg tests to match dpnp test_linalg.py style
abagusetty
committed
d924816
Fix dtype inference: use int8 trial vector so matvec preserves operator dtype
abagusetty
committed
f295bc1
add onemkl sparse gemv pybind logic
abagusetty
committed
8c68d98
sparse: add pybind11 module, CMakeLists, and hook _sparse_impl into _iterative.py
abagusetty
committed
0c4a888
Remove redundant sparse_gemv.hpp passthrough header
abagusetty
committed
4993120
sparse: gemv.cpp includes gemv.hpp directly
abagusetty
committed
a7ddc1c
sparse: capture exec_q from CSR data at closure construction
abagusetty
committed
14cb5c4
sparse/gemv: add missing headers, input validation, and MKL/SYCL exception handling
abagusetty
committed
890238c
sparse/gemv: replace explicit if/else type dispatch with 2-D dispatch table
abagusetty
committed
838dfd8

Commits on Apr 7, 2026

Commits on Apr 8, 2026

minor fixes
abagusetty
committed
e1a41b3

Commits on Apr 9, 2026

Add testing
abagusetty
committed
4442530
black formatting
abagusetty
committed
ac3bed5
Merge branch 'IntelPython:master' into feature-sparse-linalg-solvers
abagusetty
authored
3e47f4a
Update CHANGELOG.md
abagusetty
committed
a4ee24f
Merge branch 'feature-sparse-linalg-solvers' of https://github.com/abagusetty/dpnp into feature-sparse-linalg-solvers
abagusetty
committed
7e81346
remove stale testing
abagusetty
committed
c330a04
Add the missing onemkl-sycl-sparse dep to conda-recipe
abagusetty
committed
0badee4
fix pre-commit issues
abagusetty
committed
4be3d9a

Commits on Apr 13, 2026

Try again with formatting fix
abagusetty
committed
651aaeb
Merge branch 'master' into feature-sparse-linalg-solvers
abagusetty
authored
eecf9d6

Commits on Apr 22, 2026

Merge remote-tracking branch 'origin/master' into feature-sparse-linalg-solvers
abagusetty
committed
555e547

Commits on Apr 24, 2026

Commits on Apr 28, 2026

Apply remark
vlad-perevezentsev
committed
18bb358
Merge master into fix_inplace_indexind_4d
vlad-perevezentsev
committed
e64845c

Commits on Apr 29, 2026

Commits on May 6, 2026

Commits on May 7, 2026

fix the cyclic importing issue
abagusetty
committed
181eb59

Commits on May 8, 2026

Merge branch 'master' into feature-sparse-linalg-solvers
abagusetty
authored
f38ae7b

Commits on May 9, 2026

Commits on May 20, 2026

Fix pylint failures
abagusetty
committed
a375760
Merge branch 'master' into feature-sparse-linalg-solvers
abagusetty
authored
36042c6

Commits on May 26, 2026

Update dpnp/backend/extensions/sparse/CMakeLists.txt

abagusetty
and
antonwolfy
authored
902eeff
Update dpnp/backend/extensions/sparse/gemv.cpp

abagusetty
and
antonwolfy
authored
2b2ac0d
Update dpnp/backend/extensions/sparse/gemv.cpp

abagusetty
and
antonwolfy
authored
04ad7be
Update dpnp/backend/extensions/sparse/types_matrix.hpp

abagusetty
and
antonwolfy
authored
d77e516
Update dpnp/backend/extensions/sparse/gemv.cpp

abagusetty
and
antonwolfy
authored
d9734d8
Merge branch 'IntelPython:master' into feature-sparse-linalg-solvers
abagusetty
authored
8cb5656

Commits on May 27, 2026

Merge branch 'IntelPython:master' into feature-sparse-linalg-solvers
abagusetty
authored
efe896c
add left out files
abagusetty
committed
c3aae1d
Merge branch 'feature-sparse-linalg-solvers' of https://github.com/abagusetty/dpnp into feature-sparse-linalg-solvers
abagusetty
committed
ec72540
sparse: route csr_matrix.dot through oneMKL sparse::gemv
Abhishek Bagusetty
committed
2c12c20
sparse/linalg: fix dtype inference, info contract, implicit-sync bugs; add cupyx tests
Abhishek Bagusetty
committed
deea67b
sparse/linalg: fuse GMRES Arnoldi via alpha/beta gemv; one-sync CG; host Hessenberg lstsq
Abhishek Bagusetty
committed
50e9df4
sparse/linalg: opt LinearOperator out of NumPy ufunc dispatch; harden __del__ against shutdown races
Abhishek Bagusetty
committed
b395046
test: dtype-aware rtol in gmres convergence tests
Abhishek Bagusetty
committed
70ab579
gmres: use oneMKL conjugate-transpose directly; fix complex Arnoldi math bug
Abhishek Bagusetty
committed
fb11515

Commits on May 28, 2026

_interface: enforce dpnp's no-implicit-host-transfer rule
Abhishek Bagusetty
committed
e0cb7de
_interface: drop module-level numpy import; use .conjugate() and .dtype on host scalars
Abhishek Bagusetty
committed
16f7450
_interface: drop redundant numpy mentions from comments
Abhishek Bagusetty
committed
f910092

Commits on May 29, 2026

_iterative: use math.* for host scalar arithmetic; numpy stays for finfo/dtype/lstsq
Abhishek Bagusetty
committed
cfabd12
_iterative: drop dead stores and xtype alias
Abhishek Bagusetty
committed
6b2b645
sparse: address pylint findings from pre-commit
Abhishek Bagusetty
committed
7f49318
sparse: apply isort fixes
Abhishek Bagusetty
committed
70e3c30
Fix formatting
abagusetty
committed
ec1797a
test_linalg: align style with sibling cupyx tests
Abhishek Bagusetty
committed
7b05eef
Merge branch 'feature-sparse-linalg-solvers' of https://github.com/abagusetty/dpnp into feature-sparse-linalg-solvers
Abhishek Bagusetty
committed
2b43800
Merge branch 'master' into feature-sparse-linalg-solvers
abagusetty
authored
0182d6a
Move msg from v20 to v21
Abhishek Bagusetty
committed
04e4876

Commits on Jun 1, 2026

Merge branch 'master' into feature-sparse-linalg-solvers
abagusetty
authored
c4e12e6
test_scipy_sparse_linalg: fix dtype-inference test for int8 trial vector
Abhishek Bagusetty
committed
33719c1
test_scipy_sparse_linalg: aslinearoperator(numpy_array) must raise
Abhishek Bagusetty
committed
664ca0a

Commits on Jun 10, 2026

Merge branch 'master' into feature-sparse-linalg-solvers
abagusetty
authored
25cb335
Fix 2025 to 2026
Abhishek Bagusetty
committed
8eba925
sparse/types_matrix: drop dead #if branch with unused DPCTL_HAS_TYPE_DEFINED_ENTRY
Abhishek Bagusetty
committed
50f22f6
test_scipy_sparse_linalg: broader coverage, sibling-style alignment, GMRES xfail cleanup
Abhishek Bagusetty
committed
9a3f60c
Apply suggestion from @antonwolfy

abagusetty
and
antonwolfy
authored
d40b26c
test_scipy_sparse_linalg: narrow edge-case parametrizations to avoid solver-internal failures
Abhishek Bagusetty
committed
d7bd00d
test_scipy_sparse_linalg: loosen wide_spectrum residual bound for MINRES
Abhishek Bagusetty
committed
943575f
test_scipy_sparse_linalg: assert wide_spectrum against per-solver criterion
Abhishek Bagusetty
committed
2f62140

Commits on Jun 11, 2026

Merge branch 'IntelPython:master' into feature-sparse-linalg-solvers
abagusetty
authored
b8bb53b

Commits on Jun 12, 2026

Merge branch 'master' into feature-sparse-linalg-solvers
abagusetty
authored
f235123

Commits on Jun 18, 2026

Commits on Jun 19, 2026

fix onemath apis
Abhishek Bagusetty
committed
9ced832
Merge branch 'master' into feature-sparse-linalg-solvers
abagusetty
authored
0754529

Commits on Jun 23, 2026

Merge branch 'master' into feature-sparse-linalg-solvers
abagusetty
authored
7bc9852

Commits on Jul 2, 2026

Merge branch 'master' into feature-sparse-linalg-solvers
abagusetty
authored
e4e3b69
update cmake
abagusetty
committed
45a9e02

Commits on Jul 4, 2026

cleanup
abagusetty
committed
b7cfd84
Added a few more tests based on the reviews
abagusetty
committed
ec1bb1f
Merge branch 'master' into feature-sparse-linalg-solvers
abagusetty
authored
b7dc740
few more comments
abagusetty
committed
eaf7c2b
Merge branch 'feature-sparse-linalg-solvers' of https://github.com/abagusetty/dpnp into feature-sparse-linalg-solvers
abagusetty
committed
d8e2940

Commits on Jul 6, 2026

Update dpnp/backend/extensions/blas/gemv.cpp

abagusetty
and
antonwolfy
authored
90f2228
address the comments
abagusetty
committed
04cdba6
Merge branch 'master' into feature-sparse-linalg-solvers
abagusetty
authored
e8b301b
fix build
abagusetty
committed
a0ae7c2
Merge branch 'feature-sparse-linalg-solvers' of https://github.com/abagusetty/dpnp into feature-sparse-linalg-solvers
abagusetty
committed
1ace5d2
add sum_duplicates for csr
abagusetty
committed
a44c218

Commits on Jul 7, 2026

minor cleanup
abagusetty
committed
e367f9e
format files
abagusetty
committed
4074478
Revert "format files"
abagusetty
committed
598e11b
fix formatting issues
abagusetty
committed
1c7d63d
Merge branch 'feature-sparse-linalg-solvers' of https://github.com/abagusetty/dpnp into feature-sparse-linalg-solvers
abagusetty
committed
bcf05be

Commits on Jul 27, 2026

Commits on Jul 28, 2026

sparse: import dpnp.tensor (not dpctl.tensor) to fix ModuleNotFoundError
abagusetty
committed
e4647e3

Commits on Jul 29, 2026

Commits on Jul 30, 2026

Commits on Aug 3, 2026

Commits on Aug 4, 2026

Merge remote-tracking branch 'origin/feature-sparse-linalg-solvers' into feature-sparse-linalg-solvers
abagusetty
committed
f2170fa

Commits on Aug 17, 2026

Commits on Aug 18, 2026

Merge remote-tracking branch 'upstream/master' into feature-sparse-linalg-solvers
abagusetty
committed
e1585e1

Commits on Aug 24, 2026

Fix GIL for fft extensions
abagusetty
committed
f34f4ca

Commits on Aug 25, 2026

Merge remote-tracking branch 'upstream/master' into feature-sparse-linalg-solvers
abagusetty
committed
513b755

Commits on Aug 28, 2026

Commits on Aug 31, 2026

Update dpnp/tests/test_scipy_sparse_linalg.py

abagusetty
and
antonwolfy
authored
3432c89

Commits on Sep 10, 2026

Merge branch 'master' into feature-sparse-linalg-solvers
abagusetty
authored
d94c45f