ENH: vectorize loo function for local_constant estimator

[shilinng]

• closes KernelReg performance in a for loop #4187 (dealt with performance issues for leave one out estimation of bandwidth for local constant estimator)
• tests passed.
• code/documentation is well formatted.
• properly formatted commit message. See
  NumPy's guide.

This is a performance improvement for bandwidth estimation using leave one out cross-validation for local constant estimators. A method cv_loo_fast is added which is only used when reg_type is 'lc'. The idea is to calculate the leave-on-out estimations by using matrix multiplication instead of using a loop. In general, I see a 2~2.5 times of speed up for <=1000 rows of data. When using only continuous gaussian kernel, the speed up is more than 6 times due to a further vectorization in calculating the product kernel function.

Notes:

• It is essential that you add a test when making code changes. Tests are not
  needed for doc changes.
• When adding a new function, test values should usually be verified in another package (e.g., R/SAS/Stata).
• When fixing a bug, you must add a test that would produce the bug in main and
  then show that it is fixed with the new code.
• New code additions must be well formatted. Changes should pass flake8. If on Linux or OSX, you can
  verify you changes are well formatted by running

  git diff upstream/main -u -- "*.py" | flake8 --diff --isolated
  

  assuming flake8 is installed. This command is also available on Windows
  using the Windows System for Linux once flake8 is installed in the
  local Linux environment. While passing this test is not required, it is good practice and it help
  improve code quality in statsmodels.
• Docstring additions must render correctly, including escapes and LaTeX.

[pep8speaks]

Hello @shilinng! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

There are currently no PEP 8 issues detected in this Pull Request. Cheers! 🍻

Comment last updated at 2024-04-07 18:26:02 UTC