rank_score

ampligraph.evaluation.rank_score(y_true, y_pred, pos_lab=1)

Computes the rank of a triple.

The rank of a positive element against a list of negatives.

\[rank_{(s, p, o)_i}\]

Parameters:

• y_true (ndarray, shape (n)) – An array of binary labels. The array only contains one positive.

• y_pred (ndarray, shape (n)) – An array of scores, for the positive element and the n-1 negatives.

• pos_lab (int) – The value of the positive label (default = 1).

Returns:

rank – The rank of the positive element against the negatives.

Return type:

int

Example

>>> import numpy as np
>>> from ampligraph.evaluation.metrics import rank_score
>>> y_pred = np.array([.434, .65, .21, .84])
>>> y_true = np.array([0, 0, 1, 0])
>>> rank_score(y_true, y_pred)
4