更新加入 pandas - python

Question

如何在pandas原生语句（.apply()）上实现update join？我想在 Dataframe1

的 PhySci 上从 Dataframe2 更新 PSMark

Dataframe1 (tbl_ex):

|    | Sname   |   Tamil |   English |   Maths |   Science |   Sscience |   PhySci |
|---:|:--------|--------:|----------:|--------:|----------:|-----------:|---------:|
|  0 | Abu     |      35 |        65 |      64 |        98 |         36 |        0 |
|  1 | Eric    |      70 |        54 |      65 |        32 |         58 |       25 |
|  2 | Mani    |      56 |        25 |      32 |        32 |         78 |       10 |
|  3 | Ram     |      80 |        24 |      68 |        54 |         76 |        0 |
|  4 | Tom     |      40 |        26 |      56 |        69 |         42 |       65 |
|  5 | Eva     |      50 |        18 |      56 |        87 |         56 |        0 |

Dataframe2 (tbl_fy):

|    | Sname   |   PSMark |
|---:|:--------|---------:|
|  0 | Tom     |       69 |
|  1 | Ram     |       54 |
|  2 | Mani    |       32 |
|  3 | Eva     |       87 |
|  4 | Sam     |       89 |

我用 sqldf 模块实现了这个

q="""
    UPDATE tbl_ex
    SET
      PhySci = (SELECT tbl_fy.PSMark
                            FROM tbl_fy
                            WHERE tbl_fy.Sname = tbl_ex.Sname )
    WHERE
    EXISTS (
        SELECT *
        FROM tbl_fy
        WHERE tbl_fy.Sname = tbl_ex.Sname );
    """

sqldf.run(q)
print(tbl_ex.to_markdown())

tbl_ex的最终结果：

|    | Sname   |   Tamil |   English |   Maths |   Science |   Sscience |   PhySci |
|---:|:--------|--------:|----------:|--------:|----------:|-----------:|---------:|
|  0 | Abu     |      35 |        65 |      64 |        98 |         36 |        0 |
|  1 | Eric    |      70 |        54 |      65 |        32 |         58 |       25 |
|  2 | Mani    |      56 |        25 |      32 |        32 |         78 |       32 |
|  3 | Ram     |      80 |        24 |      68 |        54 |         76 |       54 |
|  4 | Tom     |      40 |        26 |      56 |        69 |         42 |       69 |
|  5 | Eva     |      50 |        18 |      56 |        87 |         56 |       87 |

Answer 1

使用，pd.merge to left merge the dataframes df1 & df2 on column Sname, Then using Series.fillna从PhySci列填充PSMark列中的NaN值，并将PSMark列分配给PhySci：

df  = pd.merge(df1, df2, on='Sname', how='left')
df = df.assign(PhySci=df.pop('PSMark').fillna(df['PhySci']).astype(int))

结果：

# print(df)

  Sname  Tamil  English  Maths  Science  Sscience  PhySci
0   Abu     35       65     64       98        36       0
1  Eric     70       54     65       32        58      25
2  Mani     56       25     32       32        78      32
3   Ram     80       24     68       54        76      54
4   Tom     40       26     56       69        42      69
5   Eva     50       18     56       87        56      87

更新加入 pandas - python

update join pandas - python

python

pandas

dataframe

sqldf