问题描述
我正在使用 Python 对时尚 mnist 数据集进行分类。我想要构建的分类器是朴素贝叶斯,在运行我构建的代码时,我下载的数据似乎具有零值,并且在某些数学计算中,变量除以这些零值,我不希望这样即将发生。所以我需要一种方法来从我想要分类的数据中“删除”这些零值。数据的示例:
[[0. 0. 0. ... 0. 0. 0. ]
[0. 0. 0. ... 0. 0. 0. ]
[0. 0. 0. ... 0. 0. 0. ]
...
[0. 0. 0. ... 0.27450982 0. 0. ]
[0. 0. 0. ... 0. 0. 0. ]
[0. 0. 0. ... 0. 0. 0. ]]
该示例仅包含 10 行 60000 条数据。
我试图用这个删除零:
Xtrain,ytrain),(Xtest,ytest) = fashion_mnist.load_data() # loading data
Xtrain = Xtrain.astype('float32') # I want from this to delete zero values
Xtest = Xtest.astype('float32') # and from that
Xtrain=Xtrain/255.0
Xtest=Xtest/255.0
#Reshaping data
Xtest=Xtest.reshape(Xtest.shape[0],Xtest.shape[1] * Xtest.shape[2])
Xtrain =Xtrain.reshape(Xtrain.shape[0],Xtrain.shape[1] * Xtrain.shape[2])
naive_bayes = NaiveBayes()
# So before i send this datas to fit function in order to run the classification i am trying
# with not working way to delete thouse values
Xtrain_non_zero = []
for i in range(len(Xtrain)):
for j in range(len(Xtrain)):
if Xtrain[i][j]!=0:
Xtrain_non_zero.append(Xtrain[i][j])
print(Xtrain_non_zero)
解决方法
暂无找到可以解决该程序问题的有效方法,小编努力寻找整理中!
如果你已经找到好的解决方法,欢迎将解决方案带上本链接一起发送给小编。
小编邮箱:dio#foxmail.com (将#修改为@)