如何在python3的keras中称呼one\u hot？

Question

我在python3中有以下代码：

docs = ['Well done!',
        'Good work',
        'Great effort',
        'nice work',
        'Excellent!',
        'Weak',
        'Poor effort!',
        'not good',
        'poor work',
        'Could have done better.']
# define class labels
labels = np.array([1,1,1,1,1,0,0,0,0,0])
from keras import backend as K


# integer encode the documents
vocab_size = 50
encoded_docs = [K.one_hot(d, vocab_size) for d in docs]
print(encoded_docs)

基本上，我尝试对列表中的每个单词进行编码，但我遇到了以下错误：

TypeError: Value passed to parameter 'indices' has DataType string not in list of allowed values: uint8, int32, int64 。你知道吗

我的代码有什么问题？这似乎是一个类型错误，但我不明白为什么。你知道吗

Answer 1

Kerasone_hot方法要求第一个参数是整数类型（在您的例子中是单词索引）。所以在使用one_hot方法之前，首先需要将每个单词映射到唯一的整数。你知道吗

docs = ['Well done!',
        'Good work',
        'Great effort',
        'nice work',
        'Excellent!',
        'Weak',
        'Poor effort!',
        'not good',
        'poor work',
        'Could have done better.']
all_words = set()
for s in docs:
    for word in s.split():
        all_words.add(word)
all_words = list(all_words)
# define class labels
labels = np.array([1,1,1,1,1,0,0,0,0,0])
from keras import backend as K


# integer encode the documents
vocab_size = len(all_words)
encoded_docs = [[K.one_hot(all_words.index(word), vocab_size) for word in d.split()] for d in docs]
print(encoded_docs)

如果要将标点符号编码为单独的单词，那么可以使用re模块拆分单词。你知道吗

import re
import string

encoded_docs = [[K.one_hot(all_words.index(word), vocab_size) for word in re.findall("[\w]+|["+string.punctuation+"]", d) for d in docs]