ENH: Moving sklearnex Ridge Regression support to oneDAL
Description
Current sklearnex Ridge uses daal4py Ridge implementation. This pr is part of moving it to onedal4py and enabling GPU support. Initial changes introduce onedal4py based Ridge to sklearnex preview.
Fixes # - issue number(s) if exists
- [x] I have reviewed my changes thoroughly before submitting this pull request.
- [x] I have commented my code, particularly in hard-to-understand areas.
- [ ] I have updated the documentation to reflect the changes, if necessary (updated in # - add PR number)
- [x] The unit tests pass successfully.
- [x] I have run it locally and tested the changes extensively.
- [x] I have resolved any merge conflicts that might occur with the base branch.
- [ ] Git commit message contains an appropriate signed-off-by string (see CONTRIBUTING.md for details)
- [ ] I have added a respective label(s) to PR if I have a permission for that.
Nice work on this so far. CI should be resolved when adding linear_model preview to setup_sklearnex script, similar to https://github.com/intel/scikit-learn-intelex/blob/main/setup_sklearnex.py#L86
Also will need to update some test files (test_memory_usage and test_monkeypatch) - you can reference this PR and do roughly the opposite: https://github.com/intel/scikit-learn-intelex/pull/1557
/intelci: run
/intelci: run
/intelci: run
Somehow TSNE is seg faulting????
/intelci: run
/intelci: run
/intelci: run
/intelci: run
/intelci: run
/intelci: run
/intelci: run
/intelci: run
/intelci: run
@DDJHB Thank you for the very good updates!
I propose to have common class for the LinearRegression and the Ridge, and have separate interface for the Ridge and LinearRegression on onedal4py side, since we would like to have SPMD iface for the Ridge as well as a next step.
/intelci: run
/intelci: run
/intelci: run
/intelci: run
/intelci: run
/intelci: run
I know its annoying, but I think the issues in CI have been sorted out. You may have to rebase this again for CI runs.
/intelci: run
/intelci: run
/intelci: run
No objections from my end at this point - just would like to see green internal CI before approval
/intelci: run
latest private ci runs: https://ecmd.jf.intel.com/commander/link/jobDetails/jobs/ef451330-b666-f193-9381-a4bf010d0e2e https://ecmd.jf.intel.com/commander/link/jobDetails/jobs/ef455961-8cad-f1bd-978b-a4bf010d0e2e https://ecmd.jf.intel.com/commander/link/jobDetails/jobs/ef4571ed-4297-f1f8-9e46-a4bf010d0e2e
@ethanglaser please review