熊猫合并语句没有给我数据帧之间的区别?

Question

似乎无法使此熊猫合并正常工作。我想知道合并前df_file_dates_from_directory和df_file_dates_from_sql之间的区别，应该20171211 - 20171218。有人可以指出我的合并语句有什么问题吗？以下是DF的打印输出

df_file_dates_from_directory
0     20171211
1     20171212
2     20171213
3     20171214
4     20171215
5     20171218
6     20171219
7     20171220
8     20171221
9     20171222
10    20171226
11    20171227
df_file_dates_from_sql BEFORE merge
File_date
0    20171219
1    20171220
2    20171221
3    20171222
4    20171226
5    20171227
#new_df  = df_file_dates_from_sql.merge( df_file_dates_from_directory['File_date'].to_frame(), indicator=True, how='outer')
new_df = pd.merge(df_file_dates_from_sql, df_file_dates_from_directory['File_date'], on='File_date', how='outer', suffixes=('_left', '_right' ) )

Answer 1

In [364]: (d1.to_frame(name='File_date')
.merge(d2, how='outer', indicator=True)
.query("_merge == 'left_only'")
.drop('_merge',1))
Out[364]:
File_date
0   20171211
1   20171212
2   20171213
3   20171214
4   20171215
5   20171218