熊猫加入具有不同索引级别/日期时间的数据框？

Question

Hi I have two DataFrames which look like this:

-------------------------------------------------
|      | dineType | menuName | unique | columns |
-------------------------------------------------
| date |          |          |        |         |
-------------------------------------------------
|%y%m%d|          |          |        |         |
-------------------------------------------------
...


-------------------------------------------------
|   | dineDate | dineType | menuName |         |
-------------------------------------------------
| 0 |  %Y%m%d  |          |          |         |
-------------------------------------------------
| 1 |          |          |          |         |
-------------------------------------------------
...

I want to join the two dataframes into one output. As you can see, the main problem is that the indexes from each table are different from each other. I want the output to follow the second table's format. Also the dates which each table starts from are different. How would I join these two dataframes?

Answer 1

如果看一下文档 ，它说您可以使用left_on ， right_on和left_index ， right_index属性基于数据框中的列和索引进行连接。

pd.merge(df1, df2, left_index=True, right_on='dineDate')

Answer 2

Instead of using string with specific format for dates, you can use pd.datetime type; after converting date and dineDate column to pd.datetime type, joining task will work without additional work. You can do that, assuming data comes from CSV file, parse_dates option of pd.read_csv . For formatting output, you can set option date_format='%Y%m%d' of pd.DataFrame.to_csv`.

Please provide sample code if you need more details.