Coverage for /pythoncovmergedfiles/medio/medio/usr/local/lib/python3.8/site-packages/keras/src/datasets/fashion_mnist.py: 35%

1# Copyright 2017 The TensorFlow Authors. All Rights Reserved. 

2# 

3# Licensed under the Apache License, Version 2.0 (the "License"); 

4# you may not use this file except in compliance with the License. 

5# You may obtain a copy of the License at 

6# 

7# http://www.apache.org/licenses/LICENSE-2.0 

8# 

9# Unless required by applicable law or agreed to in writing, software 

10# distributed under the License is distributed on an "AS IS" BASIS, 

11# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 

12# See the License for the specific language governing permissions and 

13# limitations under the License. 

14# ============================================================================== 

15"""Fashion-MNIST dataset.""" 

16 

17import gzip 

18import os 

19 

20import numpy as np 

21 

22from keras.src.utils.data_utils import get_file 

23 

24# isort: off 

25from tensorflow.python.util.tf_export import keras_export 

26 

27 

28@keras_export("keras.datasets.fashion_mnist.load_data") 

29def load_data(): 

30 """Loads the Fashion-MNIST dataset. 

31 

32 This is a dataset of 60,000 28x28 grayscale images of 10 fashion categories, 

33 along with a test set of 10,000 images. This dataset can be used as 

34 a drop-in replacement for MNIST. 

35 

36 The classes are: 

37 

38 | Label | Description | 

39 |:-----:|-------------| 

40 | 0 | T-shirt/top | 

41 | 1 | Trouser | 

42 | 2 | Pullover | 

43 | 3 | Dress | 

44 | 4 | Coat | 

45 | 5 | Sandal | 

46 | 6 | Shirt | 

47 | 7 | Sneaker | 

48 | 8 | Bag | 

49 | 9 | Ankle boot | 

50 

51 Returns: 

52 Tuple of NumPy arrays: `(x_train, y_train), (x_test, y_test)`. 

53 

54 **x_train**: uint8 NumPy array of grayscale image data with shapes 

55 `(60000, 28, 28)`, containing the training data. 

56 

57 **y_train**: uint8 NumPy array of labels (integers in range 0-9) 

58 with shape `(60000,)` for the training data. 

59 

60 **x_test**: uint8 NumPy array of grayscale image data with shapes 

61 (10000, 28, 28), containing the test data. 

62 

63 **y_test**: uint8 NumPy array of labels (integers in range 0-9) 

64 with shape `(10000,)` for the test data. 

65 

66 Example: 

67 

68 ```python 

69 (x_train, y_train), (x_test, y_test) = fashion_mnist.load_data() 

70 assert x_train.shape == (60000, 28, 28) 

71 assert x_test.shape == (10000, 28, 28) 

72 assert y_train.shape == (60000,) 

73 assert y_test.shape == (10000,) 

74 ``` 

75 

76 License: 

77 The copyright for Fashion-MNIST is held by Zalando SE. 

78 Fashion-MNIST is licensed under the [MIT license]( 

79 https://github.com/zalandoresearch/fashion-mnist/blob/master/LICENSE). 

80 

81 """ 

82 dirname = os.path.join("datasets", "fashion-mnist") 

83 base = "https://storage.googleapis.com/tensorflow/tf-keras-datasets/" 

84 files = [ 

85 "train-labels-idx1-ubyte.gz", 

86 "train-images-idx3-ubyte.gz", 

87 "t10k-labels-idx1-ubyte.gz", 

88 "t10k-images-idx3-ubyte.gz", 

89 ] 

90 

91 paths = [] 

92 for fname in files: 

93 paths.append(get_file(fname, origin=base + fname, cache_subdir=dirname)) 

94 

95 with gzip.open(paths[0], "rb") as lbpath: 

96 y_train = np.frombuffer(lbpath.read(), np.uint8, offset=8) 

97 

98 with gzip.open(paths[1], "rb") as imgpath: 

99 x_train = np.frombuffer(imgpath.read(), np.uint8, offset=16).reshape( 

100 len(y_train), 28, 28 

101 ) 

102 

103 with gzip.open(paths[2], "rb") as lbpath: 

104 y_test = np.frombuffer(lbpath.read(), np.uint8, offset=8) 

105 

106 with gzip.open(paths[3], "rb") as imgpath: 

107 x_test = np.frombuffer(imgpath.read(), np.uint8, offset=16).reshape( 

108 len(y_test), 28, 28 

109 ) 

110 

111 return (x_train, y_train), (x_test, y_test) 

112