服务器之家

服务器之家 > 正文

python MNIST手写识别数据调用API的方法

时间:2021-03-26 00:57     来源/作者:caichao08

MNIST数据集比较小,一般入门机器学习都会采用这个数据集来训练

下载地址:yann.lecun.com/exdb/mnist/

有4个有用的文件:
train-images-idx3-ubyte: training set images
train-labels-idx1-ubyte: training set labels
t10k-images-idx3-ubyte: test set images
t10k-labels-idx1-ubyte: test set labels

The training set contains 60000 examples, and the test set 10000 examples. 数据集存储是用binary file存储的,黑白图片。

下面给出load数据集的代码:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
import os
import struct
import numpy as np
import matplotlib.pyplot as plt
 
def load_mnist():
  '''
  Load mnist data
  http://yann.lecun.com/exdb/mnist/
 
  60000 training examples
  10000 test sets
 
  Arguments:
    kind: 'train' or 'test', string charater input with a default value 'train'
 
  Return:
    xxx_images: n*m array, n is the sample count, m is the feature number which is 28*28
    xxx_labels: class labels for each image, (0-9)
  '''
 
  root_path = '/home/cc/deep_learning/data_sets/mnist'
 
  train_labels_path = os.path.join(root_path, 'train-labels.idx1-ubyte')
  train_images_path = os.path.join(root_path, 'train-images.idx3-ubyte')
 
  test_labels_path = os.path.join(root_path, 't10k-labels.idx1-ubyte')
  test_images_path = os.path.join(root_path, 't10k-images.idx3-ubyte')
 
  with open(train_labels_path, 'rb') as lpath:
    # '>' denotes bigedian
    # 'I' denotes unsigned char
    magic, n = struct.unpack('>II', lpath.read(8))
    #loaded = np.fromfile(lpath, dtype = np.uint8)
    train_labels = np.fromfile(lpath, dtype = np.uint8).astype(np.float)
 
  with open(train_images_path, 'rb') as ipath:
    magic, num, rows, cols = struct.unpack('>IIII', ipath.read(16))
    loaded = np.fromfile(train_images_path, dtype = np.uint8)
    # images start from the 16th bytes
    train_images = loaded[16:].reshape(len(train_labels), 784).astype(np.float)
 
  with open(test_labels_path, 'rb') as lpath:
    # '>' denotes bigedian
    # 'I' denotes unsigned char
    magic, n = struct.unpack('>II', lpath.read(8))
    #loaded = np.fromfile(lpath, dtype = np.uint8)
    test_labels = np.fromfile(lpath, dtype = np.uint8).astype(np.float)
 
  with open(test_images_path, 'rb') as ipath:
    magic, num, rows, cols = struct.unpack('>IIII', ipath.read(16))
    loaded = np.fromfile(test_images_path, dtype = np.uint8)
    # images start from the 16th bytes
    test_images = loaded[16:].reshape(len(test_labels), 784
 
  return train_images, train_labels, test_images, test_labels

再看看图片集是什么样的:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
def test_mnist_data():
  '''
  Just to check the data
 
  Argument:
    none
 
  Return:
    none
  '''
  train_images, train_labels, test_images, test_labels = load_mnist()
  fig, ax = plt.subplots(nrows = 2, ncols = 5, sharex = True, sharey = True)
  ax =ax.flatten()
  for i in range(10):
    img = train_images[i][:].reshape(28, 28)
    ax[i].imshow(img, cmap = 'Greys', interpolation = 'nearest')
    print('corresponding labels = %d' %train_labels[i])
 
if __name__ == '__main__':
  test_mnist_data()

跑出的结果如下:

python MNIST手写识别数据调用API的方法

以上就是本文的全部内容,希望对大家的学习有所帮助,也希望大家多多支持服务器之家。

原文链接:https://blog.csdn.net/caichao08/article/details/78988389

相关文章

热门资讯

2020微信伤感网名听哭了 让对方看到心疼的伤感网名大全
2020微信伤感网名听哭了 让对方看到心疼的伤感网名大全 2019-12-26
yue是什么意思 网络流行语yue了是什么梗
yue是什么意思 网络流行语yue了是什么梗 2020-10-11
背刺什么意思 网络词语背刺是什么梗
背刺什么意思 网络词语背刺是什么梗 2020-05-22
Intellij idea2020永久破解,亲测可用!!!
Intellij idea2020永久破解,亲测可用!!! 2020-07-29
苹果12mini价格表官网报价 iPhone12mini全版本价格汇总
苹果12mini价格表官网报价 iPhone12mini全版本价格汇总 2020-11-13
返回顶部