`test_xgboost.py` test fails with big amount of cores

[amyskov]

System information

• OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Ubuntu 20.04
• Modin version: 72ca672f3e79cb4cdd589692db2d0e19d696da5c
• Python version: 3.8.12
• Code we can use to reproduce:

  conda env create --name modin -f environment-dev.yml && conda activate modin
  MODIN_ENGINE=ray MODIN_MEMORY=1000000000 pytest modin/experimental/xgboost/test/test_xgboost.py::test_xgb_with_regression_datasets[load_diabetes-112-DataFrame]

  test works fine on small machines (my laptop or Github Action machine with two cores), but fails on machine with big amount of cores (tested on Intel(R) Xeon(R) Platinum 8276L).

  Source code / logs

  E               AssertionError:
  E               Not equal to tolerance rtol=0.0007, atol=0
  E
  E               Mismatched elements: 1 / 1 (100%)
  E               Max absolute difference: 0.13829
  E               Max relative difference: 0.00085634
  E                x: array(161.62738)
  E                y: array(161.48909)
  
  modin/experimental/xgboost/test/test_xgboost.py:255: AssertionError

[Garra1980]

cc @prutskov

[vnlitvinov]

Could someone (probably @anmyachev or @YarShev) with access to multi-core machine check if this is still relevant please?

[YarShev]

Checked on 112 cores, the test passed so closing the issue.