Current sampling for random method divides the dataset only into training and validation, but for
random sampling while maintaining proportional consistency, it would have a function for the sampling to include splitting dataset to three parts: training, validation and test.
For the advanced sampling (avoiding potentially bias from random sampling), it would be ideal to split the data by the Maximum Dissimilarity Algorithm (MDA) proposed by Camus et al. (2011) : Camus, P., Mendez, F.J., Medina, R., Cofiño, A.S., 2011. Analysis of clustering and selection algorithms for the study of multivariate wave climate. Coast Engineering, 58 (6), 453–462.
