Merge 0.13.1dev3 from master to gold
#1676
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…rage` (#1654) * implement dpnp.nanmean, dpnp.nanstd, dpnp.nansum and dpnp.average * change size huge test case * fix TestArithmetic test * fix test nanmean * address comments * using assert_dtype_allclose * add rtol to comparison * Update tests/test_mathematical.py --------- Co-authored-by: Anton <[email protected]>
* Add a new impl of dpnp.linalg._lu_factor * Get dev_info_array after calling getrf * Add an extra dev_info return to _lu_factor * qwe * Add a logic for a.ndim > 2 in _lu_factor * Add an implementation of dpnp.linalg.slogdet * Add a new test_norms.py file in cupy tests * Expand test scope in public CI * A small update _lu_factor func * Remove w/a for dpnp.count_nonzero in slogdet * getrf returns pair of events and uses dpctl.utils.keep_args_alive * Update dpnp.linalg.det using slogdet * Add new cupy tests for dpnp.linalg.det * Add ipiv_vecs and dev_info_vecs in _lu_factor for the batch case * Skip test_det on CPU due to bug in MKL * Small update of cupy tests in test_norms.py * Add support of complex dtype for dpnp.diagonal and update test_diagonal * lu_factor func returns the result of LU decomposition as c-contiguous and add explanatory comments * Add getrf_batch MKL extension * Update docstring for slogdet * Add more tests * Remove accidentally added file * Modify sign parameter calculation * Remove the old backend implementation of dpnp_det * qwe * Keep lexographical order * Add dpnp_slogdet to dpnp_utils_linalg * Move _lu_factor above * A minor update * A minor changes for _lu_factor * Remove trash files * Use getrf_batch only on CPU * Update tests for dpnp.linalg.slogdet * Address remarks * Add _real_type func * Add test_det in test_usm_type * Add more checks in getrf and getf_batch functions * Improve error handler in getrf_impl * Improve error handler in getrf_batch_impl * dev_info is allocated as zeros * Remove skipif for singular tests * Implement _lu_factor logic with dev_info as a python list * Update getrf_rf error handler with mkl_lapack::batch_error * Remove passing n parameter to _getrf * Add a new test_slogdet_singular_matrix_3D test * Update tests for dpnp.linalg.det * Use is_exception_caught flag in getrf and getrf_batch error handler * Update gesv error handler * Reshape results after calling getrf_batch * Add a new dpnp.linalg.det impl and refresh dpnp_utils_linalg * Remove Limitations from dpnp_det and dpnp_slogdet docstings * Address remarks * Remove det_dtype variable and use the abs val of diag for det * Expand cupy tests for dpnp.linalg.det() * Update TestDet and TestSlogdet
* Add new blas extension and update matmul impl * Add support for N-D array add N-dimension * support more special cases + add new tests * fix random behavior on cpu * correct dtypes + support more keywords * add strided support * check input arrays * address comments - first round * address comments - second round * address comments - third round * fix pre-commit * improve test coverage * address comments * update _gemm_res_dtype func * fix a test for result_type * fix minor issues * skip tests for matmul --------- Co-authored-by: Vahid Tavanashad <[email protected]> Co-authored-by: vtavana <[email protected]> Co-authored-by: Anton <[email protected]>
* update copyright date for 2024 * update doc/conf.py * update dpnp/backend/cmake/Modules/TBBConfig.cmake
* Update dpnp.indices function * Clean up dpnp_indices * address comments * address comments * cancel changes for test_indices_list3 * Added description of all parameters * Implement dpnp.mgrid and dpnp.ogrid function (#1633) * Implement dpnp.mgrid and dpnp.ogrid function * address comments * address comments * Skip dtype check for Iris Xe * Update test_ranges.py * Added description of all parameters * Resolve pre-commit issue --------- Co-authored-by: Anton <[email protected]> Co-authored-by: Anton Volkov <[email protected]>
…tl` (#1662) * update tests update more tests rename varibale update test hypot * including both type of erros in tests * address comments
* Add a new impl of dpnp.linalg.cholesky * Add cupy tests for dpnp.linalg.cholesky * Add a batch impl of dpnp.linalg.cholesky * Remove an old impl of dpnp_cholesky * Remove DPNP_FN_CHOLESKY_EXT in dpnp_iface_fptr * Remove dpnp_cholesky_ext_c * Add a new _dpnp_cholesky_batch func * Update test_cholesky in test_sycl_queue * Expand test scope in public CI * Add more tests for dpnp.linalg.cholesky * Remove TODOs in cholesky() and update docstings * Use _common_type in dpnp_cholesky * Update dpnp_cholesky and dpnp_cholesky_batch * Keep the lexicographic order * Remove passing n parameter to _potrf * Add additional checks to potrf and potrf_batch * Extend potrf error handler * Extend potrf_batch error handler * Update tests for dpnp.linalg.cholesky * Update license year * Update cholesky docstrings * Add support upper paramenetr for potrf * Add support upper paramenetr for potrf_batch and update dpnp_cholesky * Add tests for upper parameter of dpnp.linalg.cholesky * Address remarks * Fix validation check --------- Co-authored-by: vtavana <[email protected]>
Co-authored-by: vtavana <[email protected]>
* implement dpnp.reciprocal and dpnp.angle * fix docstring * skip umath test for reciprocal * address comments * add a note for reciprocal * Update dpnp/dpnp_iface_trigonometric.py Co-authored-by: Anton <[email protected]> * fix docstring * update retunrs for angle and reciprocal update out keyword implementation for angle and reciprocal to make it consistent with implementation of other functions * Updated TestReciprocal to be compliant with the latest available dpctl in internal CI --------- Co-authored-by: Anton <[email protected]> Co-authored-by: Anton Volkov <[email protected]>
* implement sort and argsort * add more tests * update for zero dimensional arrays * address comments * fix typo
* Allow for newer version of DPC++ compiler * Stepped dpctl versions in all yaml files
* Impl dpnp.linalg.inv for 2d array * Remove an old impl of dpnp_inv * Add batch implementation of dpnp.linalg.inv func * Add cupy tests for dpnp.linalg.inf * Add dpnp tests for dpnp.linalg.inv * Add check_lapack_dev_info func * Add dev_info size check for getri_batch and getrf_batch * Add size check dev_info and error_matrices_ids * Remove dpnp_inv_ext_c * Rename check_lapack_dev_info to _check_lapack_dev_info * Skip test_inv in TestInvInvalid --------- Co-authored-by: Anton <[email protected]>
antonwolfy
changed the title
0.13.1dev2 from master to gold0.13.1dev3 from master to gold
|
View rendered docs @ https://intelpython.github.io/dpnp/index.html |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Merge the latest content from master branch to gold/2021 as part of
0.13.1dev3tag which includes changes from:dpnp.nanmean,dpnp.nanstd,dpnp.nansum, anddpnp.average#1654dpnp.reciprocalanddpnp.angle#1650outkeyword for elementwise functions #1656TypeErrortoValueErrorindpctl#1662