K-Fold Cross-Validation Splits

EasyML Coding

Problem statement

Implement a function that generates train and test index splits for K-fold cross-validation. Divide the dataset indices into k folds and return one (train, test) pair per fold.

Input

  • n_samples: total number of samples (integer).
  • k: number of folds (default 5).
  • shuffle: whether to shuffle the indices before splitting (default True).

Output

  • A list of k tuples (train_indices, test_indices), each a Python list of integers.

Requirements

  • Split the indices into k roughly equal folds. If n_samples is not divisible by k, give the extra samples to the first folds (for example, 10 samples with k=3 gives fold sizes [4, 3, 3]).
  • In iteration i, fold i is the test set and all other folds, concatenated in fold order, form the training set.
  • When shuffle=True, call np.random.shuffle() once on the index array [0, 1, ..., n_samples-1] before splitting (the seed is set by the caller).
  • When shuffle=False, keep the original order [0, 1, ..., n_samples-1].

Example

Input:

k_fold_cross_validation(n_samples=10, k=5, shuffle=False)

Output:

[([2, 3, 4, 5, 6, 7, 8, 9], [0, 1]), ([0, 1, 4, 5, 6, 7, 8, 9], [2, 3]), ([0, 1, 2, 3, 6, 7, 8, 9], [4, 5]), ([0, 1, 2, 3, 4, 5, 8, 9], [6, 7]), ([0, 1, 2, 3, 4, 5, 6, 7], [8, 9])]

Reasoning: 10 samples split into 5 folds of 2. Each fold is the test set exactly once while the other 4 folds (8 samples) train the model.

Show hint

What should the test set look like if every sample must be tested exactly once across the k iterations?

Relevant roles: ML Engineer, Data Scientist, Applied Scientist and AI Engineer.

Related free problems