Split indices into train / calibration / test sets
Usage
split_idx(n, seed = 42, p_train = 0.6, p_cal = 0.2, p_test = 0.2)
Arguments
- n
Total number of samples
- seed
Random seed for reproducibility
- p_train
Fraction for training
- p_cal
Fraction for calibration (conformal prediction)
- p_test
Fraction for test
Value
Named list with integer index vectors: train, calib, test
