Python Keras | keras.utils.to_categorical()

Keras 提供了 numpy 实用程序库，它提供了对 numpy 数组执行操作的函数。使用 to_categorical() 方法，一个 numpy 数组（或）一个具有表示不同类别的整数的向量，可以转换为一个 numpy 数组（或）一个具有二进制值且列数等于类别数的矩阵数据。

语法： tf.keras.utils.to_categorical(y, num_classes=None, dtype=”float32″)

参数：

y (input vector): A vector which has integers representing various classes in the data.

num_classes: Total number of classes. If nothing is mentioned, it considers the largest number of the input vector and adds 1, to get the number of classes.
Its default value is "None".

dtype: It is the desired data type of the output values. 
By default, it's value is 'float32'.

输出：
此函数返回二进制值矩阵（“1”或“0”）。它的行数等于输入向量的长度，列数等于类的数量。

代码：将 Cifar10 数据集标签向量转换为分类数据矩阵：

Python3

# Loading the dataset
 
from keras.datasets import cifar10
(train_images, train_labels), (test_images, test_labels)= cifar10.load_data()
 
# Labels before applying the function
# Training set labels
print(train_labels)
print(train_labels.shape)
 
# Testing set labels
print(test_labels)
print(test_labels.shape)
 
# Applying the function to training set labels and testing set labels
from keras.utils import to_categorical
train_labels = to_categorical(train_labels, dtype ="uint8")
test_labels = to_categorical(test_labels, dtype ="uint8")
 
# Labels after applying the function
# Training set labels
print(train_labels)
print(train_labels.shape)
 
# Testing set labels
print(test_labels)
print(test_labels.shape)

输出：

#Labels before applying the function
#Training set labels
array([[6],
       [9],
       [9],
       ...,
       [9],
       [1],
       [1]], dtype=uint8)

#Training set labels shape
(50000, 1)

#Testing set labels
array([[3],
       [8],
       [8],
       ...,
       [5],
       [1],
       [7]], dtype=uint8)

#Testing set labels shape
(10000, 1)

#Labels after applying the function
#Training set labels
[[0 0 0 ... 0 0 0]
 [0 0 0 ... 0 0 1]
 [0 0 0 ... 0 0 1]
 ...
 [0 0 0 ... 0 0 1]
 [0 1 0 ... 0 0 0]
 [0 1 0 ... 0 0 0]]

#Training set labels shape
(50000, 10)

#Testing set labels
[[0. 0. 0. ... 0. 0. 0.]
 [0. 0. 0. ... 0. 1. 0.]
 [0. 0. 0. ... 0. 1. 0.]
 ...
 [0. 0. 0. ... 0. 0. 0.]
 [0. 1. 0. ... 0. 0. 0.]
 [0. 0. 0. ... 1. 0. 0.]]

#Testing set labels shape
(10000, 10)

代码：考虑一个包含 7 个类的输入向量。 （它的值可以从 0 到 6(n-1)）。

Python3

# Initializing Input vector
class_vector =[2, 5, 6, 1, 4, 2, 3, 2]
print(class_vector)
 
# Applying the function on input class vector
from keras.utils import to_categorical
output_matrix = to_categorical(class_vector, num_classes = 7, dtype ="int32")
 
print(output_matrix)

输出：

[[0 0 1 0 0 0 0]
 [0 0 0 0 0 1 0]
 [0 0 0 0 0 0 1]
 [0 1 0 0 0 0 0]
 [0 0 0 0 1 0 0]
 [0 0 1 0 0 0 0]
 [0 0 0 1 0 0 0]
 [0 0 1 0 0 0 0]]