GaussianNB distributed results deviate from the scikit-learn results #771

[maliesen]

• save class_count with split=None so all ranks can access it
• get the index/mask for the global x with respect to the global y, not y.larray

Due Diligence

• General:
  • title of the PR is suitable to appear in the Release Notes
• Implementation:
  • unit tests: all split configurations tested
  • unit tests: multiple dtypes tested
  • NEW unit tests: MPS tested (1 MPI process, 1 GPU)
  • benchmarks: created for new functionality
  • benchmarks: performance improved or maintained
  • documentation updated where needed

Description

Issue/s resolved: #

Changes proposed:

Type of change

Memory requirements

Performance

Does this change modify the behaviour of other functions? If so, which?

yes / no

[brownbaerchen]

Hi, @maliesen! Thanks a lot for getting started on this!

The tests in this PR are failing because some part of our pipeline has recently changed and we have yet to fix that. This has nothing to do with you and I apologize for the inconvenience. We'll try to fix this up as soon as possible.