Reference Issues/PRs

Fixes #30009.

What does this implement/fix? Explain your changes.

• This PR introduces a new implementation of the train_test_split function that enhances its functionality with the following improvements:
  • Stratified splitting: Ensures that both the training and test sets maintain the class distribution in classification tasks.
  • Balanced regression splitting: Supports balancing continuous target variables by binning values into intervals.
  • Flexible control: Allows precise control over test and training sizes, with support for specifying the random state and whether or not to shuffle the data.

Additionally, unit tests were added to cover various edge cases, including:

• Stratified splitting for classification tasks.
• Balanced splits for regression targets.
• Custom train/test sizes.
• Shuffle and random state behavior.

Any other comments?

This enhancement provides a more flexible and robust dataset splitting utility for users working with both classification and regression problems, addressing the needs discussed in issue #30009.