如何*正确*从csv的数据读入TensorFlow
发布网友
发布时间：2022-04-27 05:59
共1个回答
热心网友
时间：2022-06-27 06:57
参考代码：
# -*- coding:utf-8 -*-import tensorflow as tf
filename_queue = tf.train.string_input_procer(["file01.csv", "file02.csv"])
reader = tf.TextLineReader()
key, value = reader.read(filename_queue)
# Default values, in case of empty columns. Also specifies the type of the
# decoded result.
record_defaults = [[1], [1], [1]]
col1, col2, col3 = tf.decode_csv(value, record_defaults = record_defaults)
features = tf.stack([col1, col2])
init_op = tf.global_variables_initializer()
local_init_op = tf.local_variables_initializer() # local variables like epoch_num, batch_size 可以不初始化local
with tf.Session() as sess:
sess.run(init_op)
sess.run(local_init_op)
# Start populating the filename queue.
coord = tf.train.Coordinator()
threads = tf.train.start_queue_runners(coord=coord)
for i in range(5):
# Retrieve a single instance:
example, label = sess.run([features, col3])
print(example)
print(label)
coord.request_stop()
coord.join(threads)
