def testStringToOneHashBucketLegacyHash(self):
with self.cached_session():
input_string = array_ops.placeholder(dtypes.string)
output = string_ops.string_to_hash_bucket(input_string, 1)
result = output.eval(feed_dict={input_string: ['a', 'b', 'c']})
self.assertAllEqual([0, 0, 0], result)
def replica_fn(elem):
# Example of typical preprocessing of string to numeric feature
hashed = string_to_hash_bucket(elem['str'], 10)
# For dense string case, slice it to size of ragged int
hashed_sliced = hashed[:, :elem['size'][0]]
# Computation with both feature from string and numeric dataset output
return elem['int'] * 10 + hashed_sliced
def insert_transformed_feature(self, columns_to_tensors):
"""Handles sparse column to id conversion."""
sparse_id_values = string_ops.string_to_hash_bucket(
columns_to_tensors[self.name].values,
self.bucket_size,
name=self.name + "_lookup")
columns_to_tensors[self] = ops.SparseTensor(
columns_to_tensors[self.name].indices, sparse_id_values,
columns_to_tensors[self.name].shape)
def insert_transformed_feature(self, columns_to_tensors):
"""Handles sparse column to id conversion."""
sparse_id_values = string_ops.string_to_hash_bucket(
columns_to_tensors[self.name].values,
self.bucket_size,
name=self.name + "_lookup")
columns_to_tensors[self] = ops.SparseTensor(
columns_to_tensors[self.name].indices, sparse_id_values,
columns_to_tensors[self.name].shape)
def testStringToHashBucketsLegacyHash(self):
with self.cached_session():
input_string = array_ops.placeholder(dtypes.string)
output = string_ops.string_to_hash_bucket(input_string, 10)
result = output.eval(feed_dict={input_string: ['a', 'b', 'c']})
# Hash64('a') -> 2996632905371535868 -> mod 10 -> 8
# Hash64('b') -> 5795986006276551370 -> mod 10 -> 0
# Hash64('c') -> 14899841994519054197 -> mod 10 -> 7
self.assertAllEqual([8, 0, 7], result)
def run_dataset_implementation(self, batch_size):
num_repeats = 5
starts = []
ends = []
for _ in range(num_repeats):
ds = dataset_ops.Dataset.from_generator(
word_gen, dtypes.string, tensor_shape.TensorShape([]))
ds = ds.shuffle(batch_size * 100)
ds = ds.batch(batch_size)
num_batches = 5
ds = ds.take(num_batches)
ds = ds.prefetch(num_batches)
starts.append(time.time())
# Benchmarked code begins here.
for i in ds:
_ = string_ops.string_to_hash_bucket(i, num_buckets=2)
# Benchmarked code ends here.
ends.append(time.time())
avg_time = np.mean(np.array(ends) - np.array(starts)) / num_batches
return avg_time
def replica_fn(elem):
# Example of typical preprocessing of string to numeric feature
hashed = string_to_hash_bucket(elem['str'], 10)
return 1000 * hashed
