input_data.py
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import gzip
import os
import numpy as np
from urllib.request import urlretrieve
from urllib.parse import urlparse
import tensorflow as tf
SOURCE_URL = 'http://yann.lecun.com/exdb/mnist/'
# 检查指定的工作目录中是否存在某个文件,如果不存在,则从指定的URL下载该文件。
def maybe_download(filename, work_directory):
if not os.path.exists(work_directory):
os.makedirs(work_directory)
filepath = os.path.join(work_directory, filename)
if not os.path.exists(filepath):
filepath, _ = urlretrieve(SOURCE_URL + filename, filepath)
statinfo = os.stat(filepath)
print('Successfully downloaded', filename, statinfo.st_size, 'bytes.')
return filepath
def _read32(bytestream):
dt = np.dtype(np.uint32).newbyteorder('>')
return np.frombuffer(bytestream.read(4), dtype=dt)[0]
# 从MNIST数据集中的图像文件中提取图像,并将其转换为一个4维的NumPy数组。
def extract_images(filename):
"""Extract the images into a 4D uint8 numpy array [index, y, x, depth]."""
print('Extracting', filename)
with gzip.open(filename, 'rb') as bytestream:
magic = _read32(bytestream)
if magic != 2051:
raise ValueError(
'Invalid magic number %d in MNIST image file: %s' %
(magic, filename))
num_images = _read32(bytestream)
rows = _read32(bytestream)
cols = _read32(bytestream)
buf = bytestream.read(rows * cols * num_images)
data = np.frombuffer(buf, dtype=np.uint8)
data = data.reshape(num_images, rows, cols, 1)
return data
# 处理分类
def dense_to_one_hot(labels_dense, num_classes=10):
"""Convert class labels from scalars to one-hot vectors."""
num_labels = labels_dense.shape[0]
index_offset = np.arange(num_labels) * num_classes
labels_one_hot = np.zeros((num_labels, num_classes))
labels_one_hot.flat[index_offset + labels_dense.ravel()] = 1
return labels_one_hot
# 从MNIST数据集的标签文件中提取标签,并根据需要将其转换为独热编码(one-hot encoding)
def extract_labels(filename, one_hot=False):
"""Extract the labels into a 1D uint8 numpy array [index]."""
print('Extracting', filename)
with gzip.open(filename, 'rb') as bytestream:
magic = _read32(bytestream)
if magic != 2049:
raise ValueError(
'Invalid magic number %d in MNIST label file: %s' %
(magic, filename))
num_items = _read32(bytestream)
buf = bytestream.read(num_items)
labels = np.frombuffer(buf, dtype=np.uint8)
if one_hot:
return dense_to_one_hot(labels)
return labels
# 处理图像数据集
class DataSet(object):
def __init__(self, images, labels, fake_data=False, one_hot=False,
dtype=tf.float32):
"""Construct a DataSet.
one_hot arg is used only if fake_data is true. `dtype` can be either
`uint8` to leave the input as `[0, 255]`, or `float32` to rescale into
`[0, 1]`.
"""
dtype = tf.as_dtype(dtype).base_dtype
if dtype not in (tf.uint8, tf.float32):
raise TypeError('Invalid image dtype %r, expected uint8 or float32' %
dtype)
if fake_data:
self._num_examples = 10000
self.one_hot = one_hot
else:
assert images.shape[0] == labels.shape[0], (
'images.shape: %s labels.shape: %s' % (images.shape,
labels.shape))
self._num_examples = images.shape[0]
# Convert shape from [num examples, rows, columns, depth]
# to [num examples, rows*columns] (assuming depth == 1)
assert images.shape[3] == 1
images = images.reshape(images.shape[0],
images.shape[1] * images.shape[2])
if dtype == tf.float32:
# Convert from [0, 255] -> [0.0, 1.0].
images = images.astype(np.float32)
images = np.multiply(images, 1.0 / 255.0)
self._images = images
self._labels = labels
self._epochs_completed = 0
self._index_in_epoch = 0
@property
def images(self):
return self._images
@property
def labels(self):
return self._labels
@property
def num_examples(self):
return self._num_examples
@property
def epochs_completed(self):
return self._epochs_completed
def next_batch(self, batch_size, fake_data=False):
"""Return the next `batch_size` examples from this data set."""
if fake_data:
fake_image = [1] * 784
if self.one_hot:
fake_label = [1] + [0] * 9
else:
fake_label = 0
return [fake_image for _ in range(batch_size)], [
fake_label for _ in range(batch_size)]
start = self._index_in_epoch
self._index_in_epoch += batch_size
if self._index_in_epoch > self._num_examples:
# Finished epoch
self._epochs_completed += 1
# Shuffle the data
perm = np.arange(self._num_examples)
np.random.shuffle(perm)
self._images = self._images[perm]
self._labels = self._labels[perm]
# Start next epoch
start = 0
self._index_in_epoch = batch_size
assert batch_size <= self._num_examples
end = self._index_in_epoch
return self._images[start:end], self._labels[start:end]
def read_data_sets(train_dir, fake_data=False, one_hot=False, dtype=tf.float32):
class DataSets(object):
pass
data_sets = DataSets()
if fake_data:
def fake():
return DataSet([], [], fake_data=True, one_hot=one_hot, dtype=dtype)
data_sets.train = fake()
data_sets.validation = fake()
data_sets.test = fake()
return data_sets
TRAIN_IMAGES = 'train-images-idx3-ubyte.gz'
TRAIN_LABELS = 'train-labels-idx1-ubyte.gz'
TEST_IMAGES = 't10k-images-idx3-ubyte.gz'
TEST_LABELS = 't10k-labels-idx1-ubyte.gz'
VALIDATION_SIZE = 5000
local_file = maybe_download(TRAIN_IMAGES, train_dir)
train_images = extract_images(local_file)
local_file = maybe_download(TRAIN_LABELS, train_dir)
train_labels = extract_labels(local_file, one_hot=one_hot)
local_file = maybe_download(TEST_IMAGES, train_dir)
test_images = extract_images(local_file)
local_file = maybe_download(TEST_LABELS, train_dir)
test_labels = extract_labels(local_file, one_hot=one_hot)
validation_images = train_images[:VALIDATION_SIZE]
validation_labels = train_labels[:VALIDATION_SIZE]
train_images = train_images[VALIDATION_SIZE:]
train_labels = train_labels[VALIDATION_SIZE:]
data_sets.train = DataSet(train_images, train_labels, dtype=dtype)
data_sets.validation = DataSet(validation_images, validation_labels,
dtype=dtype)
data_sets.test = DataSet(test_images, test_labels, dtype=dtype)
return data_sets
minist_int16_test.py
# 功能:搭建简单的网络实现手写数字识别
# 网络结构:第一层卷积 激活 池化
# 第二层卷积 激活 池化
# dropout
# softmax
import input_data
import tensorflow as tf
import numpy as np
import os
# 确保记录目录存在
if not os.path.exists('./record'):
os.makedirs('./record')
mnist = input_data.read_data_sets('MNIST_data', one_hot=True)
sess = tf.compat.v1.InteractiveSession()
# 用于将张量数据记录到文件中。
def Record_Tensor(tensor, name):
print("Recording tensor " + name + " ...")
with open('./record/' + name + '.dat', 'w') as f:
array = tensor.eval(session=sess)
_Record_Array(array, name, f)
def _Record_Array(array, name, f):
if np.ndim(array) == 1:
_Record_Array1D(array, f)
elif np.ndim(array) == 2:
_Record_Array2D(array, f)
elif np.ndim(array) == 3:
_Record_Array3D(array, f)
else:
_Record_Array4D(array, f)
def _Record_Array1D(array, f):
for i in range(len(array)):
f.write(str(array[i]) + "\n")
def _Record_Array2D(array, f):
for i in range(len(array)):
for j in range(len(array[i])):
f.write(str(array[i][j]) + "\n")
def _Record_Array3D(array, f):
for i in range(len(array)):
for j in range(len(array[i])):
for k in range(len(array[i][j])):
f.write(str(array[i][j][k]) + "\n")
def _Record_Array4D(array, f):
for i in range(len(array)):
for j in range(len(array[i])):
for k in range(len(array[i][j])):
for l in range(len(array[i][j][k])):
f.write(str(array[i][j][k][l]) + "\n")
# 构建神经网络模型时定义输入层
with tf.compat.v1.name_scope('input'):
x = tf.compat.v1.placeholder("float", shape=[None, 784])
y_ = tf.compat.v1.placeholder("float", shape=[None, 10])
# 这个函数接受一个参数 shape,它指定了权重张量的形状。
def weight_variable(shape):
initial = tf.compat.v1.truncated_normal(shape, stddev=0.1)
return tf.compat.v1.Variable(initial)
# 指定偏置张量的形状
def bias_variable(shape):
initial = tf.compat.v1.constant(0.1, shape=shape)
return tf.compat.v1.Variable(initial)
# 使用 TensorFlow 1.x 版本中的 tf.nn.conv2d 函数来执行二维卷积操作
def conv2d(x, W):
return tf.compat.v1.nn.conv2d(input=x, filters=W, strides=[1, 1, 1, 1], padding='VALID')
# 最大池化操作,降低空间维度,保留特征
def max_pool_2x2(x):
return tf.compat.v1.nn.max_pool(x, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='VALID')
# First Convolutional Layer
with tf.compat.v1.name_scope('1st_CNN'):
W_conv1 = weight_variable([5, 5, 1, 16])
b_conv1 = bias_variable([16])
x_image = tf.compat.v1.reshape(x, [-1, 28, 28, 1])
h_conv1 = tf.compat.v1.nn.relu(conv2d(x_image, W_conv1) + b_conv1)
h_pool1 = max_pool_2x2(h_conv1)
# Second Convolutional Layer
with tf.compat.v1.name_scope('2rd_CNN'):
W_conv2 = weight_variable([5, 5, 16, 32])
b_conv2 = bias_variable([32])
h_conv2 = tf.compat.v1.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2)
h_pool2 = max_pool_2x2(h_conv2)
# Densely Connected Layer
with tf.compat.v1.name_scope('Densely_NN'):
W_fc1 = weight_variable([4 * 4 * 32, 128])
b_fc1 = bias_variable([128])
h_pool2_flat = tf.compat.v1.reshape(h_pool2, [-1, 4 * 4 * 32])
h_fc1 = tf.compat.v1.nn.relu(tf.compat.v1.matmul(h_pool2_flat, W_fc1) + b_fc1)
# Dropout
with tf.compat.v1.name_scope('Dropout'):
keep_prob = tf.compat.v1.placeholder("float")
h_fc1_drop = tf.compat.v1.nn.dropout(h_fc1, keep_prob)
# Readout Layer
with tf.compat.v1.name_scope('Softmax'):
W_fc2 = weight_variable([128, 10])
b_fc2 = bias_variable([10])
y_conv = tf.compat.v1.nn.softmax(tf.compat.v1.matmul(h_fc1_drop, W_fc2) + b_fc2)
with tf.compat.v1.name_scope('Loss'):
cross_entropy = -tf.compat.v1.reduce_sum(y_ * tf.compat.v1.log(y_conv))
with tf.compat.v1.name_scope('Train'):
train_step = tf.compat.v1.train.AdamOptimizer(1e-4).minimize(cross_entropy)
with tf.compat.v1.name_scope('Accuracy'):
correct_prediction = tf.compat.v1.equal(tf.compat.v1.argmax(y_conv, 1), tf.compat.v1.argmax(y_, 1))
accuracy = tf.compat.v1.reduce_mean(tf.compat.v1.cast(correct_prediction, "float"))
tf.compat.v1.global_variables_initializer().run(session=sess)
for i in range(3000):
batch = mnist.train.next_batch(50)
if i % 20 == 0:
train_accuracy = accuracy.eval(feed_dict={x: batch[0], y_: batch[1], keep_prob: 1.0}, session=sess)
print("step %d, training accuracy %g" % (i, train_accuracy))
train_step.run(feed_dict={x: batch[0], y_: batch[1], keep_prob: 0.5}, session=sess)
print("test accuracy %g" % accuracy.eval(feed_dict={x: mnist.test.images, y_: mnist.test.labels, keep_prob: 1.0}, session=sess))
Record_Tensor(W_conv1, "W_conv1")
Record_Tensor(b_conv1, "b_conv1")
Record_Tensor(W_conv2, "W_conv2")
Record_Tensor(b_conv2, "b_fc1")
Record_Tensor(W_fc1, "W_fc1")
Record_Tensor(b_fc1, "b_fc1")
Record_Tensor(W_fc2, "W_fc2")
Record_Tensor(b_fc2, "b_fc2")
sess.close()