如何保存并稍后加载我的NLTK分类器？

Question

我正在尝试对电影评论进行情绪分析。

以下代码在函数内：

classifier = nltk.classify.SklearnClassifier(LinearSVC())

classifier.train(train_set)

如何在以后使用时保存并加载它？我应该在哪里把pickle函数放在我的代码中？感谢您的时间！

Answer 1

在运行pickle后保存和读取文档时，可以使用nltk。首先，您需要导入模块。

import pickle

classifier = nltk.classify.SklearnClassifier(LinearSVC())
classifier = classifier.train(train_set)

保存：

with open("filename", "wb") as infile:
    pickle.dump(classifier, infile)

阅读：

with open("filename","rb") as outfile:
    docs = pickle.load(outfile)

Answer 2

你可以使用pickle。

写作：

import pickle as pkl
clf = nltk.classify.SklearnClassifier(LinearSVC())
clf.train(X_train,y_train)
with open('clf.pkl','wb') as myfile:
    pkl.dump(clf,myfile)

阅读：

import pickle as pkl
with open('clf.pkl','wb') as myfile:
    clf = pkl.load(myfile)

Answer 3

保存：

def save(name,data,type='.dat'):
    name+=type
    pickle.dump(data,open(name,'wb'))#will create file if doesn't exist,will update file if there is such file

加载：

def load(name):
    return pickle.load(open(name,'rb'))#will raise error if no such file exists

例：

save('myFile','foo')
print(load('myfile.dat'))#remember .dat !
#will prints 'foo'

你需要首先导入泡菜然后你可以保存任何东西