def call(self, inputs):
if len(inputs) == 3:
X, A, I = inputs
self.data_mode = "disjoint"
else:
X, A = inputs
I = tf.zeros(tf.shape(X)[:1])
self.data_mode = "single"
if K.ndim(I) == 2:
I = I[:, 0]
I = tf.cast(I, tf.int32)
A_is_sparse = K.is_sparse(A)
# Get mask
y = self.compute_scores(X, A, I)
N = K.shape(X)[-2]
indices = ops.segment_top_k(y[:, 0], I, self.ratio)
indices = tf.sort(indices) # required for ordered SparseTensors
mask = ops.indices_to_mask(indices, N)
# Multiply X and y to make layer differentiable
features = X * self.gating_op(y)
axis = (0 if len(K.int_shape(A)) == 2 else 1
) # Cannot use negative axis in tf.boolean_mask
# Reduce X
X_pooled = tf.gather(features, indices, axis=axis)
# Reduce A
if A_is_sparse:
A_pooled, _ = ops.gather_sparse_square(A, indices, mask=mask)
else:
A_pooled = tf.gather(A, indices, axis=axis)
A_pooled = tf.gather(A_pooled, indices, axis=axis + 1)
output = [X_pooled, A_pooled]
# Reduce I
if self.data_mode == "disjoint":
I_pooled = tf.gather(I, indices)
output.append(I_pooled)
if self.return_mask:
output.append(mask)
return output
def test_indices_to_mask_rank2():
indices = [[0, 2], [1, 1], [2, 1]]
mask = ops.indices_to_mask(indices, [3, 3])
expected = [[0, 0, 1], [0, 1, 0], [0, 1, 0]]
np.testing.assert_equal(mask.numpy(), expected)
def test_indices_to_mask_rank1():
indices = [1, 3, 4]
mask = ops.indices_to_mask(indices, 6)
np.testing.assert_equal(mask.numpy(), [0, 1, 0, 1, 1, 0])
