Is there .all() or .any() equivalent in python Tensorflow

Question

Trying to find a similar operation to .any(), .all() methods that will work on a tensor. Here is a scenario:

a = tf.Variable([True, False, True], dtype=tf.bool)

# this is how I do it right now
has_true = a.reduce_sum(tf.cast(a, tf.float32)) 
# this is what I'm looking for
has_true = a.any()

Currently converting my boolean tensor into int using reduce_sum to see if there are any truths in it. Is there a cleaner way to perform this operation?

Answer 1

There are tf.reduce_any and tf.reduce_all methods:

sess = tf.Session()

a = tf.Variable([True, False, True], dtype=tf.bool)
sess.run(tf.global_variables_initializer())

sess.run(tf.reduce_any(a))
# True
sess.run(tf.reduce_all(a))
# False