Feature sparse linalg solvers by abagusetty · Pull Request #2841 · IntelPython/dpnp

abagusetty · 2026-04-09T03:38:05Z

Adds support for from dpnp.scipy.sparse.linalg import LinearOperator, cg, gmres, minres
Fixes: #2831

Have you provided a meaningful PR description?
Have you added a test, reproducer or referred to an issue with a reproducer?
Have you tested your changes locally for CPU and GPU devices?
Have you made sure that new changes do not introduce compiler warnings?
Have you checked performance impact of proposed changes?
Have you added documentation for your changes, if necessary?
Have you added your changes to the changelog?

…oneMKL hooks - _interface.py: add full operator algebra (.H, .T, +, *, **, neg), _AdjointLinearOperator, _TransposedLinearOperator, _SumLinearOperator, _ProductLinearOperator, _ScaledLinearOperator, _PowerLinearOperator, IdentityOperator, MatrixLinearOperator, _AdjointMatrixOperator, _CustomLinearOperator factory dispatch; extend aslinearoperator to handle dpnp sparse and dense arrays - _iterative.py: add _make_system (dtype validation, preconditioner wiring, working dtype selection); add _make_fast_matvec CSR/oneMKL SpMV hook; fix GMRES Arnoldi inner product to single oneMKL BLAS gemv (dpnp.dot) instead of slow Python vdot loop; offload Hessenberg lstsq to numpy.linalg.lstsq (CPU, matches CuPy); fix SciPy host-fallback tol->rtol deprecation via _scipy_tol_kwarg; add preconditioner support to CG; keep MINRES as SciPy-backed stub Refs: CuPy v14.0.1 cupyx/scipy/sparse/linalg/_interface.py, cupyx/scipy/sparse/linalg/_iterative.py"

…gmres, minres Modeled after CuPy's cupyx_tests/scipy_tests/sparse_tests/test_linalg.py. Covers: - LinearOperator: shape, dtype inference, matvec/rmatvec/matmat, subclassing, __matmul__, __call__, edge cases - aslinearoperator: dense array, duck-type, identity passthrough, rmatvec from dense, invalid inputs - cg: SPD convergence, scipy reference match, x0 warm start, b_ndim=2, callback, atol, LinearOperator path, invalid inputs, non-convergence info check - gmres: diag-dominant convergence, scipy reference match, restart variants, x0, b_ndim=2, callbacks, complex systems, atol, non-convergence info check, Hilbert-matrix stress test - minres: SPD, symmetric-indefinite, scipy reference, shift parameter, non-square guard, LinearOperator path, callback - Integration: parametric (n, dtype) cross-solver tests via LinearOperator - Import smoke tests: __all__ completeness

- Use dpnp.tests.helper: assert_dtype_allclose, generate_random_numpy_array, get_all_dtypes, get_float_complex_dtypes, has_support_aspect64 - Use dpnp.tests.third_party.cupy testing harness (with_requires, etc.) - Use numpy.testing assert_allclose / assert_array_equal / assert_raises - Use dpnp.asnumpy() instead of numpy.asarray() - Use pytest parametrize ids matching existing test conventions - Use is_scipy_available() helper from tests/helper.py - Strict class-per-solver organisation matching TestCholesky / TestDet etc.

…or dtype Two bugs fixed: 1. _init_dtype() was calling dpnp.zeros(n) which defaults to float64, so a float32 matvec would upcast and return float64, making the inferred dtype wrong. Fix: use dpnp.zeros(n, dtype=dpnp.int8) as SciPy/CuPy do — any numeric matvec will promote int8 to its own dtype. 2. _CustomLinearOperator.__init__ called _init_dtype() even when an explicit dtype was already supplied, overwriting the caller's value. Fix: _init_dtype() now short-circuits when self.dtype is already set.

…iterative.py

…ption handling Align gemv.cpp with the conventions established in blas/gemm.cpp: Headers added: - ext/common.hpp (dpctl_td_ns, consistent with other extensions) - utils/memory_overlap.hpp (MemoryOverlap guard on x vs y) - utils/output_validation.hpp (CheckWritable + AmpleMemory on y) - utils/type_utils.hpp (validate_type_for_device<T> in impl) - <sstream> (needed for stringstream error_msg) Exception handling added in sparse_gemv_impl(): - try/catch(oneapi::mkl::exception) around all oneMKL sparse calls - try/catch(sycl::exception) around all oneMKL sparse calls - release_matrix_handle cleanup in the exception error path - throw std::runtime_error with descriptive message on catch Input validation added in sparse_gemv(): - ndim checks: x and y must be 1-D - queues_are_compatible() across all 5 USM arrays - MemoryOverlap()(x, y) aliasing guard - CheckWritable::throw_if_not_writable(y) - AmpleMemory::throw_if_not_ample(y, num_rows) - keep_args_alive() at function exit (was missing, returning empty event)

… table Modeled after blas/gemm.cpp (2-D table: value type x index type) and blas/gemv.cpp (dispatch vector pattern with ContigFactory + init_dispatch_table). Changes: - Add sparse/types_matrix.hpp with SparseGemvTypePairSupportFactory<Tv, Ti> encoding the 4 supported combinations: {float32,float64} x {int32,int64} - Rewrite sparse_gemv_impl() to take typeless char* pointers (matching the blas gemv_impl signature style) — type info flows through template params only, no runtime branching inside the impl - Replace the 60-line if/else val_typenum/idx_typenum chain in sparse_gemv() with a 2-D dispatch table lookup (gemv_dispatch_table[val_id][idx_id]) - Rename init_sparse_gemv_dispatch_vector -> init_sparse_gemv_dispatch_table and implement it via init_dispatch_table<> from ext/common.hpp - All validation guards and exception handling from prior commit are preserved

…se_gemv_dispatch_table Follows the rename made in gemv.cpp when the dispatch mechanism was changed from a 1-D vector to a 2-D table (value type x index type). All other declarations (sparse_gemv signature, parameters) are unchanged.

The oneMKL 2025-2 sparse BLAS API deprecated the old 8-argument set_csr_data(queue, handle, nrows, ncols, index_base, row_ptr, col_ind, values, deps) overload in favour of a new signature that takes the sparse matrix handle as `spmat` and adds an explicit `nnz` argument: set_csr_data(queue, spmat, nrows, ncols, nnz, index_base, row_ptr, col_ind, values, deps) Fixes: - Replace old set_csr_data call with the new nnz-aware signature - Silences the resulting -Wunused-parameter warning on `nnz` (now used) - No functional change; all other logic is unchanged

…tring Line 477: `hasattr(A, "rmatmat\")` had a Markdown-escaped backslash leaked into the Python source, causing an unterminated string literal. Fixed to `hasattr(A, "rmatmat")`.

dpnp.ndarray blocks implicit NumPy conversion via __array__ to prevent silent dtype=object arrays. All test assertions must use .asnumpy() to materialize device arrays onto the host explicitly. Also replaces numpy.asarray(x_dp) in _rel_residual helper.

…dation order - _iterative.py: raise NotImplementedError for M != None *before* the _HOST_N_THRESHOLD SciPy fast-path in cg() and gmres(), so the contract is enforced regardless of system size (fixes test_cg_preconditioner_unsupported_raises, test_gmres_preconditioner_unsupported_raises). - _iterative.py: validate callback_type and raise NotImplementedError for 'pr_norm' *before* the _HOST_N_THRESHOLD branch in gmres(), so small-n systems also see the error (fixes test_gmres_callback_type_pr_norm_raises). - _iterative.py: pass callback_type='legacy' to scipy.sparse.linalg.gmres when delegating on the fast path to suppress SciPy DeprecationWarning. - test_scipy_sparse_linalg.py: add dtype=numpy.float64 to expected arange() calls in test_identity_operator and test_gmres_happy_breakdown so strict NumPy 2.0 dtype-equality checks pass (float64 result vs int64 expected).

… port SciPy corner cases

…ecurrence

- Replace .asnumpy() method calls with dpnp.asnumpy() module fn (asnumpy is not an ndarray method in dpnp; it is a top-level fn) - Fix dpnp.any(x) ambiguous truth value in x0 zero-check; replace with explicit `x0 is not None` guard for r0 initialisation - Fix V_mat.T.conj() -> dpnp.conj(V_mat.T) in GMRES Arnoldi step - Guard minres beta sqrt against tiny negative floats: sqrt(abs(...)) - Unify GMRES Hessenberg h_np assignment to avoid .real stripping producing wrong dtype for complex systems - Fix float() cast on dpnp scalar norm inside GMRES inner h_j1 line

…failures) The committed code used hypot(gbar, oldb) as delta_k which is the gamma (norm) from the PREVIOUS rotation step, not the correct diagonal entry from applying the previous Givens rotation to the current column. The correct Paige-Saunders (1975) two-rotation recurrence is: oldeps = epsln delta = cs * dbar + sn * alpha # apply previous rotation gbar_k = sn * dbar - cs * alpha # residual -> new rotation input epsln = sn * beta dbar = -cs * beta gamma = hypot(gbar_k, beta) # NEW rotation eliminates beta cs = gbar_k / gamma sn = beta / gamma w_new = (v - oldeps*w - delta*w2) / gamma # three-term update This matches scipy.sparse.linalg.minres and Choi (2006) eq. 6.11. The buggy recurrence produced solutions ~1.08x away from the true solution (rel_err ~1e0) instead of the expected ~1e-13. Co-authored-by: fix-minres-recurrence

coveralls · 2026-04-09T12:56:30Z

coverage: 77.679% (-0.8%) from 78.441% — abagusetty:feature-sparse-linalg-solvers into IntelPython:master

…lg-solvers

…feature-sparse-linalg-solvers

…agusetty/dpnp into feature-sparse-linalg-solvers

abagusetty · 2026-05-06T19:10:46Z

@antonwolfy Apologies for the delay. I got side tracked for a bit. The current workflow is failing with a sycl-post-link issue with 2026.0.0 compiler version during the link of libdpnp_backend_c.so and I see an other workflows are also affected by this.

Is this a known limitation and being triaged by compiler team

abagusetty · 2026-05-06T19:13:00Z

Seems like it is: #2887

antonwolfy · 2026-05-07T03:19:49Z

@abagusetty, yep, it's expected for the coverage due to known compiler bug, we will implement the w/a in #2887
Other workflows should pass now.

…agusetty/dpnp into feature-sparse-linalg-solvers

antonwolfy · 2026-05-11T22:47:46Z

@@ -0,0 +1,119 @@
+# -*- coding: utf-8 -*-
+# *****************************************************************************
+# Copyright (c) 2025, Intel Corporation


need to bump copyright year for new added file, should be the current year

Suggested change

# Copyright (c) 2025, Intel Corporation

# Copyright (c) 2026, Intel Corporation

antonwolfy · 2026-05-11T22:49:08Z

+
+pybind11_add_module(${python_module_name} MODULE ${_module_src})
+add_sycl_to_target(TARGET ${python_module_name} SOURCES ${_module_src})
+


Suggested change

# Ensure Cython modules build first so _usmarray.h exists

add_dependencies(${python_module_name} _usmarray)

antonwolfy · 2026-05-11T22:50:49Z

+// utils extension header
+#include "ext/common.hpp"
+
+// dpctl tensor headers


Suggested change

// dpctl tensor headers

// dpnp tensor headers

antonwolfy · 2026-05-11T22:52:21Z

+    std::int64_t, // num_rows
+    std::int64_t, // num_cols
+    std::int64_t, // nnz


Suggested change

std::int64_t, // num_rows

std::int64_t, // num_cols

std::int64_t, // nnz

const std::int64_t, // num_rows

const std::int64_t, // num_cols

const std::int64_t, // nnz

antonwolfy · 2026-05-11T22:55:09Z

+// dpctl namespace alias for type dispatch utilities
+namespace dpctl_td_ns = dpnp::tensor::type_dispatch;


Suggested change

// dpctl namespace alias for type dispatch utilities

namespace dpctl_td_ns = dpnp::tensor::type_dispatch;

// namespace for operations with types

namespace dpnp_td_ns = dpnp::tensor::type_dispatch;

antonwolfy · 2026-05-11T22:55:40Z

+#include <cstdint>
+#include <type_traits>
+
+// dpctl tensor headers


Suggested change

// dpctl tensor headers

// dpnp tensor headers

antonwolfy · 2026-05-11T23:07:49Z

+    double,       // alpha (cast to Tv inside)
+    const char *, // x (typeless)
+    double,       // beta  (cast to Tv inside)


Suggested change

double, // alpha (cast to Tv inside)

const char *, // x (typeless)

double, // beta (cast to Tv inside)

const double, // alpha (cast to Tv inside)

const char *, // x (typeless)

const double, // beta (cast to Tv inside)

antonwolfy · 2026-05-24T10:11:23Z

 * Added implementation of `dpnp.scipy.linalg.lu` (SciPy-compatible) [#2787](https://github.com/IntelPython/dpnp/pull/2787)
 * Added support for ndarray subclassing via `dpnp.ndarray.view` method with `type` parameter [#2815](https://github.com/IntelPython/dpnp/pull/2815)
 * Migrated tensor implementation from `dpctl.tensor` into `dpnp.tensor`, making `dpnp` the primary owner of the Array API-compliant tensor layer [#2856](https://github.com/IntelPython/dpnp/pull/2856)
+* Added implementation of `dpnp.scipy.sparse.linalg import LinearOperator, cg, gmres, minres` [#2841](https://github.com/IntelPython/dpnp/pull/2841)


need to move to 0.21.0 release

antonwolfy · 2026-05-24T10:42:01Z

+        # Lazy import: dpnp.scipy.sparse may import this module during
+        # package initialisation, so we cannot import it at module scope.
+        # pylint: disable=import-outside-toplevel
+        from dpnp.scipy.sparse import issparse


It seems issparse is missed and not implemented currently

antonwolfy · 2026-05-24T10:59:13Z

+# Helpers for constructing SPD, diagonally dominant, and symmetric
+# indefinite test matrices. Kept small and local, matching the style of
+# vvsort() at the top of test_linalg.py.
+def _spd_matrix(n, dtype):


Can we add CuPy tests to dpnp/tests/third_party/cupyx/scipy_tests/sparse_tests/test_linalg.py to check the compatibly?

abagusetty and others added 30 commits April 2, 2026 12:52

Add sparse.linalg iterative solvers

a7c8d31

Add sparse.linalg iterative solvers

e3e9052

Add sparse.linalg iterative solvers

c0e9fb5

Add sparse.linalg iterative solvers

943c52f

Add sparse.linalg iterative solvers

1191f7e

Fix deprecated tol kwarg in SciPy host fallback

2384185

Add tests for scipy.sparse.linalg: LinearOperator, cg, gmres, minres

58cc44b

Fix implicit numpy conversion; use .asnumpy() for dpnp arrays

3a50062

add onemkl sparse gemv pybind logic

8c68d98

sparse: add pybind11 module, CMakeLists, and hook _sparse_impl into _…

0c4a888

…iterative.py

Remove redundant sparse_gemv.hpp passthrough header

4993120

sparse: gemv.cpp includes gemv.hpp directly

a7ddc1c

sparse: capture exec_q from CSR data at closure construction

14cb5c4

minor cleanup for sparse extensions

ed58333

Fix SyntaxError: remove stray backslash in aslinearoperator hasattr s…

0a32b57

…tring Line 477: `hasattr(A, "rmatmat\")` had a Markdown-escaped backslash leaked into the Python source, causing an unterminated string literal. Fixed to `hasattr(A, "rmatmat")`.

sparse/linalg: pure-GPU CG/GMRES/MINRES, drop all CPU fallback paths,…

4292518

… port SciPy corner cases

Fix dtype.char AttributeError on dpnp dtype objects in CG/GMRES/MINRES

125dab5

Fix M guard, MINRES betacheck decay and gbar init in Paige-Saunders r…

2d753cf

…ecurrence

antonwolfy reviewed Apr 9, 2026

View reviewed changes

Comment thread dpnp/backend/extensions/sparse/gemv.cpp

abagusetty and others added 5 commits April 9, 2026 14:44

Add the missing onemkl-sycl-sparse dep to conda-recipe

0badee4

fix pre-commit issues

4be3d9a

Try again with formatting fix

651aaeb

Merge branch 'master' into feature-sparse-linalg-solvers

eecf9d6

Merge remote-tracking branch 'origin/master' into feature-sparse-lina…

555e547

…lg-solvers

antonwolfy added this to the 0.21.0 release milestone Apr 22, 2026

vlad-perevezentsev and others added 12 commits April 24, 2026 05:17

Remove __bool__ fallback from _is_boolean in _slicing.pxi

8d0149e

Support range/list as advanced index keys in dpnp_array

672a45c

Add tests for range/list advanced indexing

7ef44b7

Update changelog

5f5b0b1

Handle list/empty list advanced indexing correctly

830ef88

Apply remark

18bb358

Merge master into fix_inplace_indexind_4d

e64845c

Merge master into fix_inplace_indexind_4d

87388b4

Merge remote-tracking branch 'upstream/fix_inplace_indexind_4d' into …

9605678

…feature-sparse-linalg-solvers

Merge branch 'master' into feature-sparse-linalg-solvers

f66f7eb

Merge branch 'feature-sparse-linalg-solvers' of https://github.com/ab…

d04ca68

…agusetty/dpnp into feature-sparse-linalg-solvers

fix clang-formagt

3ab930c

abagusetty and others added 6 commits May 7, 2026 17:03

fix the cyclic importing issue

181eb59

Merge branch 'master' into feature-sparse-linalg-solvers

f38ae7b

One more try to fix the git actions

a524637

Merge branch 'feature-sparse-linalg-solvers' of https://github.com/ab…

83f6a3d

…agusetty/dpnp into feature-sparse-linalg-solvers

Fix pylint failures

a375760

Merge branch 'master' into feature-sparse-linalg-solvers

36042c6

antonwolfy reviewed May 24, 2026

View reviewed changes

	# Copyright (c) 2025, Intel Corporation
	# Copyright (c) 2026, Intel Corporation


		pybind11_add_module(${python_module_name} MODULE ${_module_src})
		add_sycl_to_target(TARGET ${python_module_name} SOURCES ${_module_src})



	# Ensure Cython modules build first so _usmarray.h exists
	add_dependencies(${python_module_name} _usmarray)

		// dpctl namespace alias for type dispatch utilities
		namespace dpctl_td_ns = dpnp::tensor::type_dispatch;

Conversation

abagusetty commented Apr 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

coveralls commented Apr 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

abagusetty commented May 6, 2026

Uh oh!

abagusetty commented May 6, 2026

Uh oh!

antonwolfy commented May 7, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

abagusetty commented Apr 9, 2026 •

edited

Loading

coveralls commented Apr 9, 2026 •

edited

Loading