Add RepeatedStratifiedGroupKFold

Describe the workflow you want to enable

Building off conversation #13621 and work already done in #18649, I'd like to add an implementation of RepeatedStratifiedGroupKFold.

Describe your proposed solution

See the implementation in #24227. Then RepeatedStratifiedGroupKFold could be used similar to below:

  >>> import numpy as np
  >>> from sklearn.model_selection import RepeatedStratifiedGroupKFold
  >>> X = np.random.randn(10, 1)
  >>> y = np.array([0, 0, 0, 0, 0, 1, 1, 1, 1, 1])
  >>> groups = np.array([1, 1, 2, 2, 2, 3, 4, 4, 5, 5])
  >>> rsgkf = RepeatedStratifiedGroupKFold(n_splits=3, n_repeats=2, random_state=42)
  >>> for train_idxs, test_idxs in rsgkf.split(X, y, groups):
  ...     # print the group assignment for the train/test indices
  ...     print("TRAIN:", groups[train_idxs], "TEST:", groups[test_idxs])
  ...     X_train, X_test = X[train_idxs], X[test_idxs]
  ...     y_train, y_test = y[train_idxs], y[test_idxs]
  TRAIN: [2 2 2 4 4 5 5] TEST: [1 1 3]
  TRAIN: [1 1 3 4 4 5 5] TEST: [2 2 2]
  TRAIN: [1 1 2 2 2 3] TEST: [4 4 5 5]
  TRAIN: [1 1 4 4 5 5] TEST: [2 2 2 3]
  TRAIN: [2 2 2 3 4 4 5 5] TEST: [1 1]
  TRAIN: [1 1 2 2 2 3] TEST: [4 4 5 5]

Describe alternatives you've considered, if relevant

No response

Additional context

No response

Start by reviewing the proposed implementation in issue #24227 and the related discussions in #13621 and #18649. Check how RepeatedStratifiedGroupKFold should expose split(X, y, groups) with n_splits, n_repeats, and random_state, and consider the supplied example output as the completion behavior.