面对 AttributeError：'list' 对象没有属性 'lower'

Posted 2023-03-12

【中文标题】面对 AttributeError：\'list\' 对象没有属性 \'lower\'

【英文标题】：Facing AttributeError: 'list' object has no attribute 'lower'面对 AttributeError：'list' 对象没有属性 'lower'

【发布时间】：2019-02-14 13:15:54

【问题描述】：

我已经发布了我的示例火车数据以及测试数据以及我的代码。我正在尝试使用朴素贝叶斯算法来训练模型。

但是，在评论中，我得到了列表列表。所以，我认为我的代码因以下错误而失败：

return lambda x: strip_accents(x.lower())
AttributeError: 'list' object has no attribute 'lower'

你们中的任何人都可以帮我解决一下我是 python 新手的问题吗....

train.txt：

review,label
Colors & clarity is superb,positive
Sadly the picture is not nearly as clear or bright as my 40 inch Samsung,negative

test.txt：

review,label
The picture is clear and beautiful,positive
Picture is not clear,negative

我的代码：

from sklearn.preprocessing import MultiLabelBinarizer
from sklearn.model_selection import train_test_split
from sklearn.naive_bayes import BernoulliNB
from sklearn.metrics import confusion_matrix
from sklearn.feature_extraction.text import CountVectorizer

def load_data(filename):

    reviews = list()
    labels = list()
    with open(filename) as file:
        file.readline()
        for line in file:
            line = line.strip().split(',')
            labels.append(line[1])
            reviews.append(line[0].split())


    return reviews, labels

X_train, y_train = load_data('/Users/7000015504/Desktop/Sep_10/sample_train.csv')
X_test, y_test = load_data('/Users/7000015504/Desktop/Sep_10/sample_test.csv')


clf = CountVectorizer()
X_train_one_hot =  clf.fit(X_train)
X_test_one_hot = clf.transform(X_test)

bnbc = BernoulliNB(binarize=None)
bnbc.fit(X_train_one_hot, y_train)

score = bnbc.score(X_test_one_hot, y_test)
print("score of Naive Bayes algo is :" , score)

【问题讨论】：

lower() 不是list 的属性。尝试将其转换为 numpy 数组。然后.lower() 应该可以工作了。

你的代码中return lambda x: strip_accents(x.lower())在哪里？

这是一段简单的代码。可以自己管理，？？？？

【参考方案1】：

我对您的代码进行了一些修改。下面发布的一个有效；我添加了有关如何调试您在上面发布的 cmets 的内容。

# These three will not used, do not import them
# from sklearn.preprocessing import MultiLabelBinarizer 
# from sklearn.model_selection import train_test_split 
# from sklearn.metrics import confusion_matrix

# This performs the classification task that you want with your input data in the format provided
from sklearn.naive_bayes import MultinomialNB 

from sklearn.feature_extraction.text import CountVectorizer

def load_data(filename):
    """ This function works, but you have to modify the second-to-last line from
    reviews.append(line[0].split()) to reviews.append(line[0]).
    CountVectorizer will perform the splits by itself as it sees fit, trust him :)"""
    reviews = list()
    labels = list()
    with open(filename) as file:
        file.readline()
        for line in file:
            line = line.strip().split(',')
            labels.append(line[1])
            reviews.append(line[0])

    return reviews, labels

X_train, y_train = load_data('train.txt')
X_test, y_test = load_data('test.txt')

vec = CountVectorizer() 
# Notice: clf means classifier, not vectorizer. 
# While it is syntactically correct, it's bad practice to give misleading names to your objects. 
# Replace "clf" with "vec" or something similar.

# Important! you called only the fit method, but did not transform the data 
# afterwards. The fit method does not return the transformed data by itself. You 
# either have to call .fit() and then .transform() on your training data, or just fit_transform() once.

X_train_transformed =  vec.fit_transform(X_train) 

X_test_transformed = vec.transform(X_test)

clf= MultinomialNB()
clf.fit(X_train_transformed, y_train)

score = clf.score(X_test_transformed, y_test)
print("score of Naive Bayes algo is :" , score)

这段代码的输出是：

score of Naive Bayes algo is : 0.5

【讨论】：

嗨@Daniel R。我们如何计算精度和召回率？

Mhmh，这种格式的数据似乎允许从 sklearn.metrics 计算混淆矩阵，但既不包括precision_score，也不包括recall_score。我会研究一下，但现在您可以打印混淆矩阵并从那里手动计算它们，方法是添加：y_pred = clf.predict(X_test_transformed)from sklearn.metrics import confusion_matrixprint(confusion_matrix(y_test,y_pred))

编辑：我成功了。必须将参数pos_label='positive' 传递给precision_score 函数，如下所示：y_pred = clf.predict(X_test_transformed) from sklearn.metrics import precision_score print(precision_score(y_test,y_pred,pos_label='positive'))

R，我已经尝试使用您上面编辑的代码。但我遇到了错误：“选择另一个平均设置”。 % y_type) ValueError: 目标是多类但平均='二进制'。请选择其他平均设置。

所以，我通过将属性 pos_label='positive' 替换为 average='micro' 来更改precision_score 和recall_score。 print("Precision Score : ",precision_score(y_test,y_pred,average='micro')) print("Recall Score :" ,recall_score(y_test, y_pred, average='micro') )

【参考方案2】：

您需要遍历列表中的每个元素。

for item in list():
      item = item.lower()

注意：仅当您遍历字符串列表 (dtype = str) 时才适用。

【讨论】：

你能在这里编辑我的代码吗？因为，我是python的新手。评论包含列表列表