Groupby 数据 - 显示最小和最大日期以及相应的值

Question

我有以下框架：

frame=pd.DataFrame(data=(['London','2020-08-31',3,4],['London','2020-09-30',1,5],
                         ['London','2020-10-31',2,6],['London','2020-11-30',7,3],
                         ['Paris','2020-08-31',9,3],['Paris','2020-09-30',1,8],
                         ['Paris','2020-10-31',4,5],['Paris','2020-11-30',2,9],
                         ['Paris','2020-12-31',3,8]),columns=["Location","Date","X","Y"])

看起来像这样：

    Location    Date    X   Y
0   London  2020-08-31  3   4
1   London  2020-09-30  1   5
2   London  2020-10-31  2   6
3   London  2020-11-30  7   3
4   Paris   2020-08-31  9   3
5   Paris   2020-09-30  1   8
6   Paris   2020-10-31  4   5
7   Paris   2020-11-30  2   9
8   Paris   2020-12-31  3   8

我想按位置分组并显示最小和最大日期及其对应的值。预期的输出是这样的：

                                                   X                               Y    
Location    Min_Date    Max_Date    Min_Date_Value  Max_Date_Value  Min_Date_Value  Max_Date_Value
London      8/31/2020   11/30/2020        3               7                4              3
Paris       8/31/2020   12/31/2020        9               3                3              8

我可以通过frame.groupby(by="Location")["Date"].agg(["min","max"]) 生成最小和最大日期，但之后我也很难将每个日期对应值的列可视化。

Answer 1

如果您想要聚合日期然后拉取相应的行，您可以尝试以下操作：

u = frame.groupby(by="Location")["Date"].agg(["min","max"])
out = (u.stack().reset_index(name='Date').merge(frame)
       .set_index(['Location','level_1']).unstack())
out.columns = out.columns.map('_'.join)

---

print(out)

            Date_max    Date_min  X_max  X_min  Y_max  Y_min
Location                                                    
London    2020-11-30  2020-08-31      7      3      3      4
Paris     2020-12-31  2020-08-31      3      9      8      3