def test_ndcg(self):
with tf.Graph().as_default():
labels = [[1., 4., 1., 0.], [4., 2., 0., 3.], [0., 0., 0., 0.]]
ranks = [[1, 2, 3, 4], [1, 3, 4, 2], [1, 2, 3, 4]]
perm_mat = [[[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 1, 0], [0, 0, 0, 1]],
[[1, 0, 0, 0], [0, 0, 0, 1], [0, 1, 0, 0], [0, 0, 1, 0]],
[[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 1, 0], [0, 0, 0, 1]]]
ndcg_ = [[0.679685], [0.95176], [0.]]
ndcg_rank = [[0.679685], [1.], [0.]]
ndcg_perm = [[0.679685], [1.], [0.]]
with tf.compat.v1.Session() as sess:
ndcg = sess.run(utils.ndcg(labels))
self.assertAllClose(ndcg, ndcg_)
ndcg = sess.run(utils.ndcg(labels, ranks))
self.assertAllClose(ndcg, ndcg_rank)
ndcg = sess.run(utils.ndcg(labels, perm_mat=perm_mat))
self.assertAllClose(ndcg, ndcg_perm)
def compute_unreduced_loss(self, labels, logits):
"""See `_RankingLoss`."""
alpha = self._params.get('alpha', 10.0)
is_valid = utils.is_label_valid(labels)
labels = tf.compat.v1.where(is_valid, labels, tf.zeros_like(labels))
logits = tf.compat.v1.where(
is_valid, logits, -1e3 * tf.ones_like(logits) +
tf.reduce_min(input_tensor=logits, axis=-1, keepdims=True))
label_sum = tf.reduce_sum(input_tensor=labels, axis=1, keepdims=True)
nonzero_mask = tf.greater(tf.reshape(label_sum, [-1]), 0.0)
labels = tf.compat.v1.where(nonzero_mask, labels,
_EPSILON * tf.ones_like(labels))
ranks = utils.approx_ranks(logits, alpha=alpha)
return -utils.ndcg(labels, ranks), tf.reshape(
tf.cast(nonzero_mask, dtype=tf.float32), [-1, 1])