最近踩坑

很久没写就会忘记
sparse_categorical_crossentropy 和 categorical_crossentropy
前者输出直接为类别 [1 2 3 0 1 2 3]
后者输出为向量[[0100][0010][0001][1000][0100][0010][0001]]

如何按行打乱dataframe
：https://blog.csdn.net/qq_22238533/article/details/70917102

from sklearn.utils import shuffle
df = shuffle(df)

两个dataframe合并：

dfA = dfA.append(dfB)

删除dataframe中值为特定值的某一行：

[https://blog.csdn.net/luocheng7430/article/details/80330566](https://blog.csdn.net/luocheng7430/article/details/80330566)

df1=df1[~df1['A'].isin([1])]