熊猫多列比较

Question

我有一个带有数据框的列的列表： List of Columns and Values

Df = ["correct_New", "miss_New", "wrong_New", "correct_Old","miss_Old","wrong_Old"]
ListNew = ["correct_New", "miss_New", "wrong_New"]
ListOld = ["correct_Old","miss_Old","wrong_Old"]

#Dropping equivalent
df_merge_files  = Df[ListNew] != Df[ListOld]

#Going to save only the differences. 
df_merge_files.to_csv("comparison_report.csv", sep=';', index=False)

ValueError：只能比较标记相同的DataFrame对象

Answer 1

尝试：

df_merge_files  = Df[ListNew].rename(columns=dict(zip(ListNew, ListOld))) != Df[ListOld]

进行这种比较时，列名必须匹配。

或者，您也可以将它们都转换为numpy并从那里工作：

df_merge_files  = Df[ListNew].to_numpy() != Df[ListOld].to_numpy()