quantificationlib.metrics.ordinal module

Score functions and loss functions for ordinal quantification problems

emd(p_true, p_pred)

Return the EMD distance between two sets of prevalences

Parameters:

• p_true (array-like, shape (n_classes, 1)) –

• p_pred (array-like, shape (n_classes, 1)) –

Returns:

emd – The EMD distance between p_true and p_pred

Return type:

float

emd_distances(A, B)

Return the EMD distances between the rows of A and B

Parameters:

• A (array-like, shape (n_samples_1, n_features)) –

• B (array-like, shape (n_samples_2, n_features)) –

• columns (A and B should have the same number of) –

Returns:

distances

Return type:

array, shape (n_samples_1, n_samples_2)

emd_score(p_true, p_pred)

Scoring metric based on EMD distance

Parameters:

• p_true (array-like, shape (n_classes, 1)) –

• p_pred (array-like, shape (n_classes, 1)) –

Returns:

emds – The EMD score

Return type:

float

References

Castaño, A., González, P., González, J. A., & Del Coz, J. J. (2022). Matching Distributions Algorithms Based on the Earth Mover’s Distance for Ordinal Quantification. IEEE Transactions on Neural Networks and Learning Systems.