def square_error(estimated, target):
with tf.name_scope('evaluation'):
with tf.control_dependencies([tf.assert_equal(count(tf.to_int32(target) - tf.to_int32(target)), 0.)]):
tf.assert_equal(count(tf.cast(target - estimated, tf.int32)), 0.)
squared_difference = tf.pow(estimated - target, 2, name='squared_difference')
square_error = tf.reduce_sum(squared_difference, name='summing_square_errors')
square_error = tf.to_float(square_error)
return square_error
def square_error(estimated, target):
with tf.name_scope('evaluation'):
with tf.control_dependencies([tf.assert_equal(count(tf.to_int32(target) - tf.to_int32(target)), 0.)]):
tf.assert_equal(count(tf.cast(target - estimated, tf.int32)), 0.)
squared_difference = tf.pow(estimated - target, 2, name='squared_difference')
square_error = tf.reduce_sum(squared_difference, name='summing_square_errors')
square_error = tf.to_float(square_error)
return square_error
def mini_batch_rmse(estimated, target):
with tf.name_scope('evaluation'):
with tf.control_dependencies([tf.assert_equal(count(tf.to_int32(target) - tf.to_int32(target)), 0.)]):
squared_difference = tf.pow(estimated - target, 2, name='squared_difference')
square_error = tf.reduce_sum(squared_difference, name='summing_square_errors')
square_error = tf.to_float(square_error)
mse = tf.truediv(square_error, count(target), name='meaning_error')
rmse = tf.sqrt(mse)
return rmse
def mini_batch_rmse(estimated, target):
with tf.name_scope('evaluation'):
with tf.control_dependencies([tf.assert_equal(count(tf.to_int32(target) - tf.to_int32(target)), 0.)]):
squared_difference = tf.pow(estimated - target, 2, name='squared_difference')
square_error = tf.reduce_sum(squared_difference, name='summing_square_errors')
square_error = tf.to_float(square_error)
mse = tf.truediv(square_error, count(target), name='meaning_error')
rmse = tf.sqrt(mse)
return rmse
def rmse(self, sess, square_error_batch, x_sparse, target, data_set, is_train): #Untested
square_error = 0
num_examples = 0
self.Train_set.reset('rmse')
data_set.reset('rmse')
steps_per_epoch = data_set.size // self.batch_size_evaluate
for step in range(steps_per_epoch):
feed_dict = self.fill_feed_dict_mini_batch(data_set=data_set,
x_sparse=x_sparse,
target=target,
is_train=is_train,
batch_size=self.batch_size_evaluate)
num_examples += sess.run(count(target), feed_dict=feed_dict)
square_error += sess.run(square_error_batch, feed_dict=feed_dict)
mean_square_error = square_error / num_examples
rmse = math.sqrt(mean_square_error)
rmse *= 2
return rmse
def rmse(self, sess, square_error_batch, x_sparse, target, data_set, is_train): #Untested
square_error = 0
num_examples = 0
self.Train_set.reset('rmse')
data_set.reset('rmse')
steps_per_epoch = data_set.size // self.batch_size_evaluate
for step in range(steps_per_epoch):
feed_dict = self.fill_feed_dict_mini_batch(data_set=data_set,
x_sparse=x_sparse,
target=target,
is_train=is_train,
batch_size=self.batch_size_evaluate)
num_examples += sess.run(count(target), feed_dict=feed_dict)
square_error += sess.run(square_error_batch, feed_dict=feed_dict)
mean_square_error = square_error / num_examples
rmse = math.sqrt(mean_square_error)
rmse *= 2
return rmse